Does Stainless Steel Jewelry Tarnish? What Stains
Home
Yiwu, Zhejiang, China
Mon – Fri 9:00 AM – 6:00 PM · Sat 9:00 AM – 4:00 PM
[email protected]

Does Stainless Steel Jewelry Tarnish? What Actually Stains

Alan Xiao, Founder of Sourcing Pioneer
Alan Xiao Founder of Sourcing Pioneer
· Updated November 16, 2026 · 11 min read

Bare stainless steel does not build the dark film that silver builds. What buyers call tarnish on a steel part is usually a coating, a joint or a film of dirt.

That difference decides who pays. A complaint that names the wrong cause leads to the wrong fix, and often to a refund you did not owe.

This guide names the six things that stain a steel part, then gives you a test that separates them in ten minutes. It ends with the four lines that stop most of the complaints before they start.

Brushed steel ring and chain on a beige studio surface, one band showing a dull grey patch
The dull patch is rarely the steel. On a coated part it is the layer above the steel, worn thin at a contact point.

The Short Answer

Stainless steel can discolour. It cannot build the film that silver builds in a drawer.

Silver needs hydrogen sulfide to go dark, and it takes that gas straight out of the air. A steel part in normal wear has no such path, because the chromium oxide skin on the surface blocks the reaction.

So a brown or black mark on steel jewellery has another source. Almost always it is one of these six.

CauseWhere you see it firstIs the steel at fault
Coating worn throughInner band, clasp edge, the high points of a chainNo
Silver solder at a jointJoints and jump rings onlyNo
Free iron from the workshopScattered specks on flat facesPartly, the surface was dirty
Chloride pit in a creviceEngraved grooves and spring pocketsYes, at that one spot
Residue filmThe whole part, an even toneNo
Clear coat going yellowRecessed areas, or the whole partNo

Only two rows on that list come from the metal itself. The other four are things sitting above it, and each one behaves differently when you try to clean it.

What The Steel Does When It Meets Air

Stainless steel is iron with enough chromium to hold oxygen away from the iron. The usual cut-off is 10.5 percent chromium.

That chromium forms a thin oxide skin, a few atoms deep. The skin rebuilds itself in air after a scratch, as long as the surface is clean.

The skin is the whole product. A scratch that stays shallow does not start rust, because the skin closes over it again within hours.

What breaks the skin is dirt, salt and a wet surface held against it. That is a short list. It is also why most complaints turn out to be about handling, not about the alloy.

The grade on the tag is the first thing to read, and it is printed far less often than it should be. Four grades cover almost every steel part in a jewellery catalogue.

Grade on the tagChromiumNickelMolybdenumWhat it means on a jewellery line
304, also sold as 18-8about 18 percentabout 8noneLowest cost steel part. It stains first near salt water.
316L16 to 18 percent10 to 122 to 3The usual jewellery grade. The molybdenum is why salt hurts it less.
43016 to 18 percentnonenoneMagnetic and cheap. A steamy bathroom shortens its life.
20116 to 18 percentlow, with manganese insteadnoneA cheaper substitute sold as stainless. Stains earlier than 304.

Those 316L figures are the ones published for the low-carbon austenitic grade, and they are the numbers a mill certificate should repeat.

Two of those numbers do the work. Nickel raises the metal cost and slows staining. Molybdenum is the one that buys tolerance to salt, and it is the reason 316L became the default for parts worn against skin.

The 18-8 label is a recipe, not a brand. It means roughly 18 percent chromium and 8 percent nickel, which is why it reads close to 304. The site has a separate note on what the 18/10 stamp promises on the kitchen side of the same family.

A magnet tells you nothing useful here. 304 and 316L are both non-magnetic in the annealed state. A deep draw or a sharp bend can make either one slightly magnetic at the edge. A supplier who offers the magnet as proof is guessing.

Rust Is Not Tarnish

Steel does have a failure mode of its own, and the trade has a plain name for it: tea staining. It is brown, it looks alarming, and it sits on the surface.

Discolouration of the surface of stainless steel that does not affect the structural integrity or the longevity of the material.

That definition comes from an industry bulletin on tea staining, and the useful part is the second half. A brown film that sits on the surface has not eaten the part.

The cause is chloride plus water that stays put. Sea salt holds water at very low humidity. A coastal surface therefore stays wet far longer than the air suggests, and it keeps reacting the whole time.

Two details make it worse. A rough finish gives the salt somewhere to sit, and a crevice stops the water from drying at all. Both matter more than grade in the first month.

The same bulletin sets the coastal band at about five kilometres from surf, and up to twenty in the right wind. It puts grade 316 as the minimum inside that band.

For jewellery, distance to the sea is a red herring. A ring lives in a harsher micro-climate than a balcony railing. Think of a wet finger inside a rubber glove, a wrist under a sleeve, or a hand in a pool.

One more line from that bulletin is worth repeating. Hydrochloric acid should never touch stainless steel, because it stains the surface and can start worse damage. That rules out a common household shortcut for cleaning metal.

What you seeWhat it isHow deepDoes it growWhat removes it
Dark film on silverSilver sulfideThe first few layers of atomsYes, while sulfur is aroundSilver dip or a silver cloth
Tea stain on steelIron oxide on the skinVery shallowYes, while salt stays wetA wash, then a fine polish
Rust on steelIron oxide into the metalInto the partYes, and it pitsNothing that keeps the part

Read that table before you open a complaint. A silver dip will do nothing for tea staining, and a polishing paste will not stop a pit.

The Six Things That Actually Stain A Steel Part

Each cause leaves a different mark, in a different place. Learn the address and you have the answer.

Six labelled cards naming the causes of marks on steel jewellery, each with the place the mark appears
Six causes, six addresses. Where the mark sits narrows the list before you test anything.

One, the coating is spent. On a plated or coated part, the colour is not the steel. When the layer thins, steel grey shows through, and the whole piece reads as tarnished when only one face has changed.

Plating and vapour deposition wear differently but end in the same place. A plated layer is soft and wears at edges. A deposited layer is harder and thins slowly across the whole face. Either way the colour is a layer, and a layer has a life.

Two, the joint is silver solder. Many steel parts are assembled with a silver-bearing solder, and silver darkens from sulfur. The joint goes first while the links beside it stay bright.

That pattern is easy to read. If the marks sit only at joints and jump rings, the steel is clean and the complaint is about assembly.

Three, the workshop left iron on the part. A polishing room that also handles carbon steel puts iron dust on everything. Those particles rust in place and leave small brown specks on a flat face.

This is the one factory fault on the list, and it is the easiest to fix. A passivation bath before packing dissolves the free iron, and the specks never appear.

Four, a pit started in a crevice. A pit needs a place where water sits. A spring pocket, an engraved groove or a gap under a setting does the job, and one brown dot becomes a rough spot.

A pit is the only mark on this list that damages the part. It is also the slowest to appear, which is why it usually shows up after the return window.

Five, something dried on it. Soap, hand cream, sunscreen and hard water all dry into a film. It dulls the whole part evenly and wipes off with a damp cloth.

A film is the most common complaint and the cheapest to answer. It is also the one most often mistaken for a quality problem, because it appears on the first wearing.

Six, the clear coat went yellow. Some parts carry a lacquer that holds a colour or a matte look. Lacquer yellows with sun and heat, and it lifts at the edges first.

The steel under a yellowed coat is unchanged. Strip the coat and the part is fine, so the answer is a coating note, not a replacement.

How To Tell Them Apart In Ten Minutes

Work from the cheapest test upwards. Most complaints stop at the second step.

Six step ladder of checks running from a damp cloth and a location check up to a passivation record
Cheapest test first. A damp cloth settles most of these before a polish is needed.

Step one, wipe it. Use a damp cloth with a drop of mild soap, then dry it. A residue film leaves, and a mark that stays is real.

Step two, find where it starts. A mark at an edge, a clasp or a chain high point is wear. Scattered specks on a flat face suggest free iron. One dot inside a groove points to a pit. An even dullness over the whole part means a film.

Step three, polish a hidden spot. Put a little metal polish on the inside of a band and rub with a soft cloth. A stain that lifts and does not return within a day was residue or surface oxide.

Step four, look at the colour underneath. Hold the part under a lamp at a low angle. Gold over steel reflects a warm tone even when it is thin. Steel grey showing through means the layer has opened.

Step five, ask for the salt spray hours. Every coated part has a number, and it is quoted against a named method. A supplier who cannot give one has no coating spec, only a colour.

Step six, ask for the passivation record. Three items: the standard used, the bath chemistry, and which acceptance test was run on the lot.

Those six steps fit on one page of a notebook. Run them once on a returned sample. You will know which of the six causes you are looking at before the customer finishes the email.

One test does not belong on the list. A magnet cannot tell 304 from 316L, and it cannot see a coating. It only answers a question nobody asked.

The Layer That Fails First

A gold-coloured steel ring is two products stacked. The steel carries the shape; the layer carries the colour. The colour is what the customer bought, and it is the part that wears out.

The layer is thin, and it is measured in microns. One micron is one thousandth of a millimetre, which is about a hundredth of the thickness of a sheet of paper.

Wear follows contact. The inside of a band, the edge of a clasp and the high points of a curb chain touch more than the rest. They go first. That is why a “tarnished” ring usually shows one bright grey patch and gold everywhere else.

The chain shape decides where the grey appears. A curb chain sits flat on its face, so the outer links take the rub from a sleeve. A box chain hides its wear inside the tube until the tube opens.

Two numbers settle a coating argument, and neither one works alone. Ask for the thickness in microns, and for the salt spray hours against a named method. A thickness alone does not predict life, because adhesion decides how fast the layer lifts.

The choice between the two coating routes belongs on that same sheet. The site’s comparison of vapour coating against a plating bath sets out where each one earns its price.

Nothing about this is a defect. Every colour over steel is a coat, and every coat thins. The failure is in selling it as permanent.

That is also why a plated part needs a different claim from a solid one. The word “plated” is not a warning; it is a description of a two-part product. The base metal underneath the finish is the part that carries the weight.

What Stops It At The Factory

The control is a chemical step, not a better alloy. It is called passivation.

Passivation does not add a coating. It removes free iron and dirt from the surface, so the chromium oxide skin can form evenly again. The steel underneath is unchanged, which is why the weight is the same after the bath.

The standard is ASTM A967. It covers three routes for passivating steel parts: immersion in nitric acid, immersion in citric acid, and an electrochemical treatment. ASTM A380 is the companion practice for cleaning and passivating.

The citric route is the one a small jewellery shop can run without a fume problem. A967 sets the bath at 4 to 10 percent citric acid, with three temperature and time pairs.

Citric treatmentBath temperatureTime in the bath
160 to 71 C4 minutes
249 to 60 C10 minutes
321 to 49 C20 minutes

The nitric route is described by A380 codes. F is 20 to 50 percent nitric acid. G adds sodium dichromate. K and L are weaker baths for the free-machining grades.

After the bath, A967 asks for a test, not a look. There are five: water immersion, high humidity, salt spray, copper sulphate, and a potassium ferricyanide-nitric acid test.

The last one is the most direct. It turns blue where free iron remains, so a pass is the absence of blue, not a clean appearance.

Two limits are worth knowing before you write the line. Pickling is not passivation: pickling uses nitric and hydrofluoric acid to take off weld scale, while passivation is nitric or citric alone. And a high-carbon grade such as 440C should not go into an acid bath at all, because it can crack.

StepWhat it removesWhat it cannot fix
DegreaseOil and polishing compoundIron embedded in the surface
PickleHeat tint and scale from a weldA pit that has already opened
PassivateFree iron from tooling and handlingA coating that has worn through

A part that leaves the shop without this step looks fine for about a month. That delay is exactly why the complaint lands on the customer’s desk instead of on the loading dock.

Ask the workshop for the bath record with the lot number, not a promise. The shop that can name its own standard is the one that will still be there at the second order.

Six Lines For The Order Sheet

A drawing that says “steel ring, gold colour” leaves every stain decision open. Six lines close them, and each one is checkable on arrival.

Line 1, the grade. 316L, written as 316L and not as “surgical steel”. Ask for the mill certificate for the coil or the casting lot that went into the order.

Line 2, the passivation. ASTM A967, citric route, with the treatment number and the acceptance test named. A bath record with the lot number belongs in the same folder.

Line 3, the coating. Type, thickness in microns, adhesion test, and salt spray hours against a named method. Four items, one line, no adjectives.

Line 4, the joint. Laser weld, silver solder or adhesive, named per joint. This is the line that explains a mark at a jump ring six months later.

Line 5, the claim. The exact care wording allowed on the card and in the listing. Approve the words, not just the idea, because the listing is what a marketplace refund is judged against.

Line 6, the retained sample. One piece from each lot, sealed and labelled with the date. A claim argued six months later is won with a sample, not with a memory.

Those six lines cost nothing to write. They also give a factory a reason to run the bath, because the line it has to sign is the line it will keep.

Answering A Tarnish Complaint

A message that says “it tarnished” is not yet a claim you can settle. Three photographs change that: the whole part, the mark close up, and the same mark after a wipe.

Then ask where it started. A mark at the clasp and a mark on a flat face lead to different answers, and one of those answers is not your problem.

Sort the reply into three buckets. A surface film is a care note. A spent coating inside its stated life is a replacement. A pit on a part sold as 316L is a material question, and it needs the certificate before you say anything.

The care card does most of the work, and four lines cover it. Take the piece off before a pool or a shower. Wipe it dry after wear. Keep it out of a steamy bathroom drawer with silver. Never use a chlorine-based cleaner on it.

One wording change cuts more claims than any coating upgrade. Write “steel with a gold layer” instead of “gold steel”, and print the thickness beside it. A buyer who knows the layer is a layer does not expect it to last forever.

That leaves the honest limit. A coated steel part will show its edge eventually, and a bare 316L part will not. If the product page promises the second and the carton holds the first, no care card will save the review.

SIX CAUSES

I see this complaint most weeks, and the photo usually shows a worn edge, not a stained alloy. Send a picture of the mark, the grade on the tag and the coating line from your spec. I will name the cause and tell you whether the answer is a care note, a replacement or a supplier conversation.

Send The Photo

Questions Buyers Ask

Does 316L stainless steel jewelry tarnish?

The surface can discolour, but it does not build the sulfide film that silver builds. Most marks come from a coating, a solder joint or a dried residue.

Will stainless steel jewelry rust in the shower?

Water alone is not the problem. Chloride and water trapped in a crevice are. Rinse the piece and dry it, and a bare 316L part will last for years.

Can I wear it in a pool or the sea?

You can, and then you should rinse it in fresh water and dry it. Chloride is what starts tea staining, and a wet surface keeps the reaction going.

How do I remove brown marks from a steel piece?

Wash it with mild soap and warm water first, then dry it. If a mark stays, use a fine metal polish on a soft cloth. Skip the baking soda paste.

Is a magnet a good way to check the grade?

No. 304 and 316L are both non-magnetic when annealed, and a bend can make either one slightly magnetic at the edge. Ask for the mill certificate instead.

What should I ask the factory for in writing?

The grade, the passivation standard with its acceptance test, and the coating thickness with its salt spray hours. Three items, all checkable, all cheap to confirm.

Questions & Comments

We read every one and reply within 24 hours

Which of the six marks did your customer photograph, and where on the part does it sit? Send the picture and the grade, and I will tell you whether the steel is involved at all.

Subscribe
Notify of
guest
0 Comments
Most Voted
Newest Oldest
0
Would love your thoughts, please comment.x
()
x