Open your own shop. Identifying a toy is free. Create an account →
Hammer & Mold Est. 2026
The Molders Sofubi Shop Past results The workshop Retrobrighting Moulds Selling Sell Packages We buy Instagram X Facebook Sign in
Know the toy. Then own it. Est. MMXXVI EU shipping, tracked
Vintage Kenner · Star Wars · Ghostbusters

Know the toy.
Then own it

Hammer & Mold brings together original vintage toys, the companies that made them and the stories that made them famous. Explore the history. Watch the commercials. Buy the originals.

Meet the Molders Browse the collection
The shop

For sale

Live auctions on original vintage toys. Place a bid, or take it straight away with Buy now.

The makers

Meet the Molders

We organize the collection the way it was built: maker first. The maker is the molder; its toy lines live beneath it. Filter the lines by their maker, or sort them across the years.

Sort
Kenner

Star Wars

1977 – 1985

The line that invented the modern 3¾-inch action figure, and sold an empty box one Christmas to do it.

Meet the Molder →
Kenner

The Real Ghostbusters

1986 – 1991

Proton packs, ecto-plasma and gross-out mutations. The strangest, most inventive toys on the shelf.

Meet the Molder →
BullmarkMarusanPopy

Kaiju

1966 – 1984

Soft-vinyl monsters from Tokyo's tokusatsu boom. Godzilla, Ultraman and the beasts of late-night television, hand-poured and hand-painted.

Meet the Molder →
MaruyamaMedicomGoto

Sofubi

1953 – now

The material itself. Hollow Japanese soft vinyl, slush-cast in heated metal moulds, pulled out by hand and sprayed by hand. And Hiroshi Goto, who has been painting it for seventy years.

Meet the Molder →
Playmates

Teenage Mutant Ninja Turtles

1988 – 1997

Mutant reptiles, pizza and ooze. The line that turned a black-and-white comic into a worldwide phenomenon.

Meet the Molder →
LJN

E.T. the Extra-Terrestrial

1982

The gentle alien of 1982. Poseable figures, glowing fingers and the most famous bicycle in cinema.

Meet the Molder →
Galoob

The A-Team

1983 – 1984

Mr. T, machine guns and a black van. No plan survived contact, but the toys did.

Meet the Molder →
Hasbro

WWF Wrestling

1990 – 1994

Four inches of hard plastic, one move each, and a ring to throw them in.

Meet the Molder →
Imperial

Godzilla & King Kong

1985 – 1986

The monsters you could actually own. Cheap vinyl on a hook by the till, from the company a Holocaust survivor built out of a ten-cent bouncing ball.

Meet the Molder →
Palitoy

Star Wars, in cardboard

1978 – 1985

Britain's Star Wars line, from the company that also made Action Man. Its Death Star is printed card because plastic cost too much, which is exactly why it is now the one collectors chase.

Meet the Molder →
Thinkway Toys

Toy Story

1995

The toys about toys. Pull-string Woody and Buzz, straight from Pixar's first feature.

Meet the Molder →
Why we exist

Not just a grid of photos

Anyone can sell an old toy. We want to tell you what you're actually looking at.

A vintage King Kong figure cranking the handle of an Agfa movie projector.
The Toy

Know what you're buying

Year, maker, series, scale, accessories and condition. The details collectors actually want to know.

The Story

Know where it came from

Who made it, when it appeared and what made it special. The story behind the plastic.

The Commercial

See how it was sold

The original TV commercials that introduced these toys to the world, right where they belong.

Home/ Retro Brighting
The workshop

Retro Brighting

Why white plastic goes yellow, and what peroxide actually does about it

A white Stormtrooper. A Ghostbusters Ecto-1. A Dreamcast. Give them enough years and they all head for the same nicotine yellow, and the forums will tell you it is the fire retardants. That is close enough to sound right and wrong enough to matter. Here is what is actually happening inside the plastic, what forty years of handling adds to it, the two recipes we use, and the step afterwards that decides whether any of it lasts.

Written by Ramon Gervais/Sourced, with the papers listed/Hammer & Mold, Haarlem
2-HMA
The compound doing the yellow. The toy built it itself
220 h
Under UV, by which half the flame retardant has left the plastic
12%
The strongest peroxide sold in the EU, and our ceiling too
10 yr
Before the treated half of a test shell went yellower than the untreated half
The rot

The rubber that saves the toy is the part that fails.

ABS is three things in one name. Acrylonitrile for hardness. Styrene for stiffness and gloss. Butadiene, which is synthetic rubber, and it is the reason a Kenner figure bends instead of shattering when it goes down the stairs.

That rubber is also the weak one. It carries double bonds, which are the spare links in a molecule where something else can attach. The other two do not have them. A double bond is an open door.

Light, or just heat over enough years, knocks a hydrogen off the butadiene and leaves behind a free radical: a molecule with an unpaired electron, desperate to grab something. Oxygen obliges within moments. Then it loops. Radical, hydroperoxide, carbonyl, another radical. Every pass shortens the chain a little and lines up a few more double bonds in a row, which chemists call conjugation.

Conjugation is colour. A run of double bonds soaks up the blue end of the light, so what comes back to your eye is yellow going on brown. The restoration world names a specific culprit, 2-hydroxymuconic acid, and it may well be right. It hardly matters. The shape of the answer is not in dispute.

The yellow is not dirt sitting on the toy. It is the toy itself changing.

Which decides everything below. You cannot wash off a colour the plastic built out of its own backbone. And the same reaction that made the colour also cut the chains that made the plastic tough, so the yellowing and the brittleness are the same ageing read two ways. Everything on this page deals with the first one. Nothing here touches the second.

The additive

The flame retardants, and what everyone gets wrong

Toys and electronics moulded in ABS through the 1970s and 80s came with brominated flame retardants in the mix, because a big hollow plastic object near a heater is a fire risk and the standards said so. They are still in there. What they do is more specific than the folklore.

CompoundWhat it isWhat it does under UV
TBBPA
Tetrabromobisphenol‑A		The dominant brominated flame retardant in ABS, and the one most studied because of electronic wasteSpeeds up the oxidation of the ABS around it, and removes the induction period: the grace window a stabilised plastic normally gets before it starts to go
DecaBDE
Decabromodiphenyl ether		Used in high‑impact polystyrene, the other white-goods plastic of the eraBreaks down fast. In dispersed HIPS film it follows first‑order kinetics at 0.14 per second, past half converted within minutes of exposure
HBCDD, DBDPE, TTBPTThe later replacements, brought in as the earlier ones were restrictedSame failure, slower. HBCDD runs at 0.019 per second. All of them shed brominated aromatics; bromophenols are the most abundant product from TTBPT
Antimony trioxideNot a retardant on its own. The helper packed alongside the bromine so less of it is neededTurns up in every analysis of flame‑retarded ABS from e‑waste. It is why a bromine reading and an antimony reading travel together
The mechanism is the same one that makes them useful. A flame retardant works because the bond holding the bromine is weak enough to break under heat, and the loose bromine then mops up the radicals feeding the fire. Ultraviolet light breaks that same bond at room temperature. The property that stops your toy burning is the property that will not leave it alone for forty years.

So here is the correction. Everybody repeats that the bromine turns the plastic yellow. It does not. The yellowing is the ABS oxidising, and it happens in plastics with no flame retardant in them at all.

What the bromine does is make it happen sooner and faster. It is not the pigment. Think of it as fuel thrown onto a fire that was already burning. That is a real effect and a significant one, and it is not the same claim.

The difference matters because it predicts what happens after you treat a toy. Under UV, roughly half the TBBPA and a fifth of the bromine leave the plastic inside 220 hours, and the broken-down bromine does not sit still. It travels through the polymer. Bleach the surface and there is a reservoir underneath, free to work its way out. That is the mechanism behind the most common complaint about retro brighting: it comes back.

The other layer

And then there are your hands

A toy that was played with carries something a warehouse toy never did. A film of sebum, the oil your skin makes, put there daily for years by a child who did not wash first. It is not cosmetic and it is not inert.

  • Squalene is the yellow one. It is twelve to fifteen percent of sebum and it carries six double bonds, which by now you know is an invitation. It oxidises more readily than the other oils around it, and it is the component identified as responsible for yellow staining on fabric. It does the same thing on a white toy. Nobody launders the toy.
  • Once it oxidises it starts feeding the reaction. Squalene does not just sit there being yellow. It breaks down into radicals, on the surface of a plastic whose whole failure mode is radical attack. The fingerprint joins in, working from the outside in.
  • The fatty acids act as solvents. They are close to sixty percent of sebum. They swell ABS slightly, wash the protective additives out of the surface layer, and crack a moulding that was left under tension. On soft PVC, a cape or a vinyl kaiju, they speed up plasticiser migration. That is the tacky, greasy figure every collector has met.
It explains something you have probably noticed. Yellowing on a played-with figure is often patchy rather than even, and sometimes the pattern reads as a grip. That is two things at once: the bulk oxidation running everywhere, and a local dose of oxidiser and solvent left exactly where a hand held it. Sometimes you can almost see where they were held.

Wash it first. Always. Lukewarm water, a drop of pH-neutral detergent, a soft brush, dried properly and not left damp in a seam. That is it.

Do that and you have removed a fuel source, not treated a symptom. It is cheap, it is safe, and you can undo it by doing nothing. Removing forty years of skin oil is cleaning. Everything after this line is intervention.

The chemistry of the fix

Peroxide bleaches. It does not repair.

Hydrogen peroxide attacks those lined-up double bonds. Break the run and it stops absorbing blue light, so your eye stops seeing yellow. Same trick as taking the colour out of a tomato stain. The molecule is still there. Its ability to show a colour is gone.

So the plastic looks younger. Chemically, it is not younger.

The UV light is not decoration. It splits the peroxide into hydroxyl radicals, which is what makes this take hours instead of geological time. The oxy laundry booster in every recipe is sodium percarbonate with TAED, which reacts with peroxide to make peracetic acid, a sharper oxidant that works at room temperature.

Peroxide plus that booster plus UV is the whole of what people mean by retro brighting. It was worked out on the English Amiga Board over the summer of 2008, after the CBM Museum Wuppertal published a five-day soak that March. Peroxide on yellowed household plastic goes back to the 1960s as a home remedy.

What it cannot do is anything structural. It does not rebuild the butadiene. It does not reverse the chain scission, the cutting of the molecular chains that made the plastic tough. It does nothing about the brittleness, which is the half of the damage that actually threatens the object.

A retrobrighted toy is a bleached toy, not a restored polymer. It is exactly as degraded as it was and now looks like it is not. Hold that in mind and every decision below gets easier.

From the bench

Our own pieces, before and after

No borrowed before-and-after here. Everything below was treated by us with the method set out further down, then protected, and shot in the same light from the same angles both times. Change the lighting between two photographs and almost any restoration looks impressive. That proves nothing. Both columns of each pair move together every three seconds, or on the arrows.

Kenner Real Ghostbusters Ecto-1 1984 · treated September 2026
Before1 / 3
A Kenner Real Ghostbusters Ecto-1 before treatment, front three-quarter from the left. The body has yellowed to a deep cream, the roof lights read olive and the windows grey-green. The same yellowed Ecto-1 from the rear three-quarter on the right, gunner seat raised with its red guns, an orange ghost figure standing on the table in front of it. The same yellowed Ecto-1 side on with the front to the right, showing the full length of the cream body against the red fin flash.

Before. Forty years of everything at the top of this page. The body has gone to a deep cream, and it did not go alone: the roof lights read olive and the glazing grey-green. Washed and dried first, then photographed against a neutral ground.

After1 / 4
The same Ecto-1 after treatment and a coat of UV protectant, front three-quarter from the left. The body is white and the roof lights have gone back to clear blue. The treated Ecto-1 from the rear three-quarter on the right: white body, blue roof lights and blue glazing, with the Ghostbusters badge on the door and the tailgate. The treated Ecto-1 front three-quarter from the left and slightly higher, with the orange ghost figure sitting on the gunner seat. The treated Ecto-1 photographed from above, showing the white roof, the four blue roof lights and the black gunner platform.

After. Treated, rinsed, then finished with 303 against the UV that caused all of it. We stopped at a match with the parts around it. The no-ghost badges are new stickers, not the originals. The number plates are original and were left alone, which is why they still carry the old tone against the new white.

The clear parts are the half nobody photographs. The roof lights and the glazing had yellowed along with the shell, and they came back with it. But the protectant is not for them: it must not go on clear plastic. So one car now has a shielded surface and an unshielded one, centimetres apart, on the same shelf, under the same light.

Which makes this car an experiment rather than an illustration. Nobody has published what happens to a treated and protected piece over ten years. This one was bleached and protected in September 2026 and carries its own control.

If the mechanism above is right, the glazing yellows first and the white holds longer. If they go at the same rate, the protectant is doing less than it claims. Either answer is worth having. The only cost is photographing this car again in a year, same angles, same light.

And we took a good sticker off on purpose. This one arrived as three states at once. The tailgate badge was already gone. One door carried a replacement somebody had fitted in the wrong size. The other door still had a sound original.

Treat that and you get a clean white car still wearing three states, which reads as a muddle rather than a history. So they all came off, the good one included, and it wears new ones. The number plates stayed, and they are the only original markings left on it.

That is the judgement the rule further down leaves you with. Never take peroxide to a sticker settles what you must not do. It does not settle what to do with stickers that were already past saving before you started. Here the answer was to make it one honest state instead of three, and then to say so out loud.

Nintendo Super Nintendo, PAL 1992 · treated September 2026
Before1 / 3
A Super Nintendo PAL console stripped to its parts and laid out flat before treatment. Every moulded piece has gone to a deep khaki, darkest on the upper shell, while the internal white cartridge guide at the lower right is still almost white. The yellowed upper shell seen at an angle with the smaller mouldings and the bare circuit board beside it. The shell is deep khaki while the controller port strips and the EXT cover next to it are noticeably paler. Close-up of the yellowed top panel with the four-colour Super Nintendo logo, the olive cast running right across the moulding.

Before. A PAL machine, model SNSP‑001A, moulded in 1992 and sold here as SNSP‑HOL. It came apart before anything else happened: the board, the shielding and the printed labels never go near the liquid. The yellowing is not even. The upper shell is the worst of it and the small mouldings beside it are visibly paler, on the same console, from the same years.

After1 / 3
The same Super Nintendo parts laid out flat after treatment. The khaki has gone and the mouldings read as cool grey, with the two-tone grey of the original design visible again between the body and the top panel. The treated mouldings at an angle: the power slide, the reset button, the EXT cover and the front badge plate, all back to grey, with the circuit board and the metal shield beside them. The console part-assembled after treatment, the grey body carrying the darker grey top panel, with the Super Nintendo badge plate and the PAL VERSION strip laid out in front of it.

After. Treated, rinsed and finished with 303. The two greys are back, and they are supposed to be there: this machine left the factory with a light body and a darker top panel, so a finish that had evened them out would have been a worse result, not a better one. The label on the underside and the serial sticker were masked and are untouched, which is why they still read the same in both columns.

The white part at the lower right of the first photograph is the cartridge guide, and it lives inside the machine. Same console, same thirty-four years, same air, no light. It is the nearest thing to a control sample a console can carry, and it is the argument for the whole page in one frame: what yellows is what was looked at.

What it is. A PAL Super Nintendo, model SNSP‑001A, with SNSP‑HOL on the plate: the Dutch machine. Japan had it as the Super Famicom from November 1990 and America from September 1991, but Europe waited. Britain got it in April 1992 and the rest of the continent that June, and the underside of this one is stamped 1992. It is a first-year console, thirty-four years old, and it was bought here.

There is nothing rare about it. Nintendo sold 49 million of them worldwide and kept the European ones in production until 1998. What is rare is a clean one, and that is entirely a story about plastic.

Why this machine in particular. The shell is ABS loaded with a brominated flame retardant, which the fire-safety rules of the period required. That bromine is the accelerant named at the top of this page: it does not start the reaction, it speeds up the one the ultraviolet begins. It is the reason a console from 1992 browns in a way a toy of the same age often does not.

Which is also why it browns unevenly. The dose was not constant. Nintendo adjusted it over the production run, and a console is assembled from mouldings that need not have come from the same batch, so two parts that spent the same decades in the same living room can sit a shade apart. Add the obvious: the top of a console faces the room and the underside faces the carpet.

Both of those are in the first photograph. The upper shell is the darkest thing in the frame, the small mouldings beside it are paler, and the cartridge guide that never saw daylight is still white.

Choosing a strength

How strong, and what it is called in the shop

Peroxide comes two ways. As a clear liquid sold by percentage, and as hairdressing cream sold by volume. Volume is how much oxygen gas it will give off, which is why the numbers look odd. They line up neatly.

3%
10 volume
6%
20 volume
9%
30 volume
12%
40 volume
Start herePharmacy. Slow enough to let you change your mind
Rarely neededBuy it only if you own neither of the other two
Heavier workWhere 12% feels like too much
The ceilingThe strongest anyone may sell you over a counter
Gentler, slower, more forgivingFaster, and a narrower window to stop
StrengthSold asWhere you find itUse
3%10 volumePharmacy, drogist. LiquidLight yellowing, small unpainted parts, and anything you are not sure about. Slow enough to forgive you
6%20 volumeHairdressing supplier. Liquid or creamThe middle setting. Rarely worth buying if you have the other two
9%30 volumeHairdressing supplier. CreamHeavier yellowing, where 12% feels like too much
12%40 volumeHairdressing supplier. Cream or liquidThe legal ceiling. Annex III of EU Regulation 1223/2009 caps hydrogen peroxide in hair products at 12%, present or released, so 40 volume is the strongest thing anyone will sell you over a counter
Stronger does not mean better. It means faster. The reaction cannot tell the difference between the colour you want gone and the surface you want kept. All a higher concentration buys you is less time to notice you have gone too far. Start at the top of that table.
Formula one

3%, submerged

This is where we start. Light yellowing, unpainted parts, anything small enough to go right under. It is also what we use when a piece is valuable enough that we would rather wait than hurry.

IngredientAmountWhy
Hydrogen peroxide 3%, 10 volume500 ml, or enough to cover the partOrdinary pharmacy peroxide is fine. An industrial grade buys you nothing here
Oxy laundry booster, sodium percarbonate with TAED¼ teaspoon, roughly 1 gram per 500 mlThe activator. Stir it in just before you use it and never store the mix with it in, because it gasses off and the bottle becomes a pressure vessel
Glycerine1 level teaspoon, optionalSlows evaporation at the surface. It matters more in the gel than it does here
  • Get it completely under, in glass or clear PET. Do not leave half of it above the liquid. The waterline turns into a treatment line you will never even out afterwards, because that band gets liquid and air at once. If it floats, weigh it down or turn it every hour.
  • UV-A or daylight, 20 to 25 degrees. A lamp at 30 to 45 cm is plenty. Do not build a greenhouse. Heat drives the same oxidation that damaged the plastic in the first place, and a hot sealed jar of peroxide is a problem of its own.
  • Then check it. And check it again. At 3% we look about once an hour. Four to twelve hours is normal for light yellowing, and a stubborn piece can take two or three days. There is no magic finishing colour. Stop when it matches the original plastic around it, not when it looks like a new refrigerator.
  • Rinse in cold water, properly. Peroxide trapped in a screw hole or a joint keeps reacting in the dark for weeks after the part looks dry. Dry it fully before anything goes back together.
Formula two

12%, as a gel

We keep this for heavier yellowing, and for parts too large or too awkward to submerge. It is much faster. That is the advantage and the problem. This is the original Retr0bright recipe, which specifies 10 to 15%.

IngredientAmountWhy
Hydrogen peroxide 12%, 40 volume500 mlOr skip the next two rows and buy 12% cream developer, which is already about the right consistency
Xanthan gum2 heaped tablespoonsTurns liquid peroxide into a gel that will stay on a vertical surface
Glycerine1 level teaspoonKeeps the gel from drying out, which is what actually causes most streaking
Oxy laundry booster¼ teaspoonIn at the moment of use, never before. Store the gel without it, in a dark container
  • Blend, rest, blend. Peroxide and xanthan for five seconds, glycerine in, five seconds more, stand a minute, then five seconds again. You want it smooth. Lumps make thin spots, thin spots dry first, and uneven drying is where streaks come from.
  • One to two millimetres, evenly, then cling film. Brush it on thicker than looks necessary and wrap the whole part. The film is not optional. Gel left open to the air dries from its edges inward, and the line between wet and drying gel is exactly where blotching appears.
  • Under UV, look every 30 minutes. Four times the strength is roughly a quarter of the comfortable window. Unwrap it. Look at it. Wrap it again if it needs more. Do not trust a timer.
  • Treat 12% with respect. Nitrile gloves and eye protection. Somewhere ventilated. Away from metal. Never mixed with other cleaners. Never heated, never sealed warm. It bleaches skin on contact and it does not warn you first.
The evidence against

Ten years later, the treated side looked worse

Anyone thinking about doing this should see this experiment first.

In 2015 part of a Dreamcast shell was masked off with tape while the rest was retrobrighted. Treated and untreated ABS then sat side by side on the same object, in the same room, under the same light, for ten years.

It was compared again in 2025. The treated areas had yellowed more than the untreated ones beside them, and carried streaking and blotching the masked areas did not. Not slower. Not equal. Worse.

That is one Dreamcast shell, not a controlled study, and it deserves to be called that. But as hobby experiments go it is unusually good, because the sample and the control are the same piece of plastic. And what happened makes chemical sense. The bleaching removed the colour near the surface. It did not remove the degraded material underneath, it did not repair the broken chains, and it did nothing to stop light, heat and oxygen carrying on.

Conservation science has no reversal to offer either. The Getty runs a plastics programme precisely because the field does not yet know how to save these objects, and nothing in it endorses bleaching them. We can take the colour off. We cannot repair the polymer. That is not going to change.

But look at what that shell did next. It went back on a shelf with nothing at all between it and the light that had yellowed it the first time. A decade of the original cause, on a surface with less left to lose.

So this is not an argument against retrobrighting. It is an argument against retrobrighting something and then putting it straight back on the same sunny shelf, which is what almost every published before-and-after actually shows.

After the bleach

The part people tend to forget

Bleaching removes the visible result of the oxidation. It does nothing to the oxidation. So once a part comes out of the gel, the only question left is what you put between it and the thing that started all of this.

Read the chain backwards and there are three things driving it. Light and heat get it started. Oxygen keeps it going. The bromine coming off the flame retardant speeds it up. Peroxide addresses none of them. It removes what they produced.

Of those three, exactly one is under your control on a shelf. You cannot take the oxygen away. You certainly cannot take the flame retardant out of something moulded in 1984. You can take the light away.

That is why we use a UV protectant afterwards. We currently use 303 Aerospace Protectant, and the reasons are as much about what is not in the bottle as what is.

303 Aerospace ProtectantWhat it isWhat it means here
The formulaWater-based, with no petroleum distillates, no alcohol, no mineral oil and no silicone oils. Dries clear and matteThose absences are the point. Petroleum distillates and silicone oils would sit on the surface doing to the plastic roughly what sebum does, which is an expensive way to undo two days of work
What it doesUV blockers, sold against discolouration, fading, embrittlement, cracking and chalkingIt targets the thing that starts the reaction. It is the only step on this whole page that addresses a cause instead of a symptom
Where it must not goNot for clear plasticsWindscreens, bubbles, canopies, the Ecto‑1 glazing, the front of a display case. On a toy the clear parts are usually the ones you least want to fog
How oftenEvery three to five weeks under sun, two to three in extreme sunThose intervals are written for a boat deck. A shelf indoors is a far gentler exposure and nobody has published a figure for it, so treat it as a coating you maintain rather than a seal you apply once
The console restoration community got here on its own: applying 303 after treatment is common practice, and the wikis that recommend it say plainly that nobody knows how well it holds up long term. That is the honest way to put it, and we are not going to put it more strongly.

What it cannot do, plainly. It does not stop thermal oxidation, which runs in the dark. It does not stop the degraded bromine already inside from migrating outward. It does not touch the brittleness.

Our expectation is simply that a protected piece should yellow more slowly than the same piece under the same light without it. That is a prediction, not a result, because the long-term data does not exist yet.

Which is why we write down what was done to every treated toy we sell, and when. In ten years that record will be worth more than any before-and-after photograph.

Will it yellow again?

Yes, eventually. Here is how to slow it down

The short answer to the question everyone asks after a retrobright.

A retrobrighted toy will yellow again. The bleach removed the colour, not the cause, and the light, heat and oxygen that yellowed it the first time are still there. The Dreamcast shell above shows what happens when nothing changes afterwards.

How fast it comes back depends almost entirely on where the toy lives. These are the things that make the difference:

  1. Keep it out of direct sunlight and away from windows. Daylight through glass still carries enough UV to do it.
  2. Apply a UV protectant afterwards, such as 303 Aerospace Protectant, and reapply it rather than treating it as a one-time seal.
  3. Keep it away from heat: radiators, spotlights, a hot attic. Heat drives the oxidation even in the dark.
  4. Handle it with clean, dry hands. The oil from skin speeds the yellowing up where you hold it most.
  5. For display, use LED lighting and a case with UV-filtering acrylic, or keep it in a closed cupboard or box.

Do all of that and a treated piece should yellow much more slowly than it did before. That is the honest promise: slower, not never.

Our position

Where we draw the line

We do retrobright toys. A good part of the shelf has been through it. But there are places we stop, and none of them are negotiable for a better-looking photograph.

  • Never over paint, tampo printing, stickers or glue. Peroxide lifts printing, dulls paint, softens adhesive and finds the edge of a sticker. Losing an original factory marking is a far bigger problem than yellow plastic, because that marking is often what dates the piece.
  • Never a multi-material toy as one job. A head, torso and arms can look identical and be different plastic from different tools. They bleach at different speeds. That is how one yellow figure becomes one white, one cream and one grey figure.
  • Never when the ageing is what identifies the object. On a bootleg, an undocumented cast or an odd variant, the surface itself is part of the evidence. Take it off and it cannot be put back, on the one kind of object where the object is the whole archive.
  • Never without saying so. Every treated toy we sell is described as treated, alongside what was used. A buyer should know whether the white they are looking at is forty years old, or forty years old and recently bleached.

And there is a fifth, which is not a rule so much as a habit. If the yellowing is even, the paint and printing are intact and the thing simply looks its age, we usually leave it alone.

Age is not automatically damage that needs fixing.

The short version

If you remember nothing else

Wash it first. Lukewarm water, pH-neutral detergent, soft brush, dried properly. That alone removes something that is still working on the plastic.

Then decide whether it needs treating at all. A great deal of vintage plastic looks perfectly good with even, honest ageing, and doing nothing is a real answer.

If you do treat it: the lowest concentration that will do the job. Check it far more often than feels necessary. Stop when it matches the original plastic around it, not when it reaches refrigerator white. Rinse it properly.

Then protect it from UV, because an unprotected bleached surface is the Dreamcast. Store it cool, dry, out of direct sun, off fluorescent tubes, and not sealed in a PVC bag.

And write down what you did. Somebody looking at that toy in ten years deserves to know what happened to it.

Where this comes from

The mechanism is from the peer-reviewed literature. The formulas are from the restoration community that worked them out, credited by name. Where the two disagree, the paper wins and the page says so.

Keep reading

The molders

1947 – 2000KennerStar Wars and The Real Ghostbusters, in the white ABS this page is about. 1947 –Marusan, Bullmark, PopySoft vinyl, which fails a different way entirely. 1986Dor MeiWhere the surface of the toy is the only archive there is. 1953 –SofubiThe vinyl itself, and the man who paints it.
Home/About
About

It started with a shelf

Hammer & Mold · Netherlands

A private collection that grew into something bigger. This is where it comes from, and what we hunt for.

What began as a handful of childhood pieces turned into a hunt: for the odd, the overlooked and the wonderfully strange. The finding, the careful cleaning, the search for a missing weapon or a lost accessory, the stories behind each toy. Somewhere along the way the collecting became a way of living.

We are drawn to the weird ones. Toys that were played with, loved and actually handled, over the pristine pieces sealed and graded never to be opened again. A toy locked in plastic, never to be read or felt, becomes a frozen artifact. We would rather keep them alive.

Have a collection of your own? We buy and take on whole collections and single pieces. Tell us what you have, send a few photos, and we will look after it.

Home/Sell
Sell with us

Upload. AI builds the listing

Photos in. Listing ready.

Upload a few clear photos. Our AI identifies the toy, dates it, checks the condition and builds the listing for you. It also reviews recent sales to suggest a fair starting price.

Identification is free. You only use a listing when you decide to sell.

How it works

From photos to listing in two steps

1 · Upload & guide

Add photos and direction

Add clear photos of the front, back, accessories and any visible flaws. Include a close-up of the maker's stamp where possible.

You can also add your own source or a note of direction, either to point us at something you already trust or to have the identification double-checked against it.

A worn or partly legible stamp is fine.

2 · Review & list

We identify, check and grade it

Our AI reads the toy the way Hammer & Mold does, dating it from the moulded stamp rather than the box, then identifies the year, line and variant, grades the condition and reviews recent sales to suggest a price.

You see the identification, the sources, the condition, the pricing and the listing details before anything goes live. Change what you like, set your Reserve or Buy Now price, then approve the listing.

Optional, and worth a minute. We can grade what we can see, but we cannot tell from a photograph whether the eyes still light up, and on a toy that has electronics that is usually the largest single thing about what it is worth. If you already know the reference, the variant or a source that settles it, put it here and we will check against it.
Your photos stay in your browser until you start identification. Identification is free. A listing is only used when you choose to publish the item for sale.
Your toy details will appear here.
Selling packages

What it costs to sell

Toy identification is free and unlimited. Photograph anything and we will identify it, date it, show what originally came with it and grade its condition. You only pay when you list. Choose a package based on how much you sell, or let us work out which one costs you least.

Home/We buy
Sell it to us

Or skip the auction and take the money today

Whole collections and single pieces

You do not have to run an auction to get rid of a box of toys. Send us photographs, we tell you what it is worth, and we make you an offer for the lot. You post one parcel and we pay. No listing, no waiting, no wondering whether it sells.

Two ways, and one of them is better for you

Sell it to us, or sell it through us

We would rather you picked the right one than the one that suits us, so here is the arithmetic on a toy we think will fetch €30.

On a toy worth about €30Sell it to usSell it through us
You receive€ 15,00€ 25,50
WhenWithin a weekAfter it sells
If it does not sellOur problemYou still have it
Photographs, listing, packingWe do itYou do it
If it turns out to be rareWe keep the upsideYou keep it
We pay about half of what we think a piece will fetch, and we say so up front. That is what buying it outright costs us: we take the risk that it does not sell, we do the photographing and grading and packing, and we wait for our money instead of you. If you would rather have the other €10,50 and do the work yourself, list it and keep 85%. Both offers are real and the page for the other one is here.
What we take

Bring the box, not the figure

The honest reason: assessing one toy and assessing forty costs us nearly the same in setup, so a whole collection is worth far more of our attention than a single common figure.

Yes, gladly
  • Whole collections, boxes, attic finds, shop stock
  • Single pieces we think will fetch €25 or more
  • Played-with condition. Paint loss and loose joints are fine
  • Incomplete figures, spares, accessories and weapons in a bag
  • Anything with a moulded maker's stamp, even a worn one
A collection has no minimum per piece. Send the lot and we price it as a lot.
Not for us
  • Single common figures under about €25
  • Modern reissues and toys still on shop shelves
  • Broken beyond display: melted, snapped, missing limbs
  • Bootlegs presented as the licensed original
  • Anything that is not a toy, model or related memorabilia
A single cheap figure is not worth what it costs either of us to move it. List it yourself, or put it in with a box of others.
How it goes

Photographs, an offer, a parcel

1. You send photographs. A few clear shots of what you have, laid out. For anything you think is worth something, add a close-up of the moulded stamp on the feet or the lower back: that text is what lets us date it and find what it really sells for.

2. We tell you what it is, before we make an offer. You get the identification either way, and you keep it whether you sell to us or not. If we think a piece is worth more than we would pay for it, we will say so and tell you to list it yourself.

3. We make one offer for the lot. Within two working days, with the reasoning attached: what we think each piece fetches and what we are paying for it. No haggling and no fishing.

4. You post it, we pay. We send you a prepaid label. Payment goes out the day the parcel arrives and matches the photographs. If something is not as described you get the parcel back at our cost, not a lower offer.

Show us

What have you got?

Photographs and a way to reach you. Nothing else needed, and nothing is committed by sending it.

We reply within two working days. Your photographs are used to make you an offer and for nothing else.
Home/Sell/Packages
Selling packages

Pick the package that fits how you sell

Toy identification is always free and unlimited. Photograph anything and we will tell you what it is, who made it, when it was made and what condition it is in. You only pay when you sell. Choose a package based on how much you sell each month. Every package includes price research, buyer protection and live postage rates.

Which one, exactly

See which package costs you least

Tell us roughly how much you sell. We will compare the monthly fee and selling fee together and recommend the cheapest package.

The number that matters

What the platform takes out of a sale

Compare what buyer and seller pay together. On a €100 hammer price, Catawiki charges the seller 12.5% and adds another 9% buyer fee. The buyer pays €109 while the seller receives €87,50. Hammer & Mold adds no buyer fee. A €100 hammer price means the buyer pays €100.

On a €100 hammer priceBuyer paysSeller receivesTotal platform fees
Catawiki€ 109,00€ 87,5021,5%
Hammer & Mold, Starter€ 100,00€ 85,0015,0%
Hammer & Mold, Trader€ 100,00€ 88,0012,0%
Hammer & Mold, Dealer€ 100,00€ 92,008,0%
Monthly package fees are not included in this table. The calculator above includes both your package fee and selling fee to show which package costs least at your sales volume.
Catawiki figures based on its published seller commission and buyer protection fee, August 2026.
Is the monthly fee worth it?

What each package really costs

A monthly package fee buys you a lower selling fee. Sell enough in a month and the saving becomes greater than the subscription. Sell less, and Starter is cheaper. The table shows the total cost of each package at different monthly sales volumes.

Changing your package

Move up. Move down. Any time.

  • Upgrades happen immediately. Your new listing allowance is available straight away. You only pay the difference for the rest of the month.
  • Downgrades start next month. Your current package stays active until the end of the period you have already paid for. You can cancel the change before then.
  • Unused listings stay with you. They roll over up to twice your monthly allowance. Downgrading never removes listings you have already accumulated.
  • No minimum term. No phone calls. Change or cancel your package from your account whenever you need to.
Home/Shop/Seller
Active lots

On auction now

Live lots from this seller.

Home/Shop/Lot
Home/ Lot/ Checkout
Delivery

Where should it go?

We need the delivery address before payment, so the postage you are charged is the postage for your own country and not an estimate.

Home/Sold archive
Results

Sold & ended

Past results. What has changed hands, and for how much.
Home/Your account