Dating vintage clothing on eBay from a photo of the tag

What a tag can date, and what it can't

You look for tag features that didn't exist before a particular year, and every one you find pushes the garment's earliest possible date forward. That gives you a floor. It almost never gives you a year.

That's where most vintage guides go wrong. They print tag features as though each one dates a garment. A tag can tell you a shirt wasn't made before 1972. Nothing on a tag can tell you the shirt wasn't made last spring.

Start with what's missing instead of what's printed, because the rules that put labels into American clothing only ever added requirements. Then read the last column before you read the dates: three rows cite a federal regulation and link to its text, two rest on union mergers with documented dates, and three are conventions the hobby settled on. Treating those groups as equally solid is how people end up confidently wrong.

The table has one blind spot, and it's the important one. Every row describes the tag. A tag is a separate object from the garment it's sewn into, and separate objects can be moved.

Care labels, fiber content and country of origin

All three tell you the earliest year a garment could have been made, because each became a labeling requirement on a date you can look up.

• Care labels. The useful one, because its absence is visible in a photograph. The rule's own exemptions section, 423.8(f), refers back to "the Care Labeling Rule issued on December 9, 1971," and the hobby's working cutoff is around July 1972, the point at which the requirement bit. Check the scope before you lean on it: 423.1 covers finished garments including hosiery but excludes footwear, gloves and hats outright, so a bare-tagged cap proves nothing.
• Fiber percentages. A tag reading 50 percent cotton and 50 percent polyester is doing what the Textile Fiber Products Identification Act rules require. Those rules carry a source note of 24 FR 4480, June 2, 1959, so a printed percentage breakdown generally means after about 1960.
• Country of origin. A "Made in USA" line is the third stackable date, and the hobby puts it after about 1971. We found no dated cutoff in the regulations we were willing to quote, and a date you can't cite is a date you shouldn't assert.

Two cautions. These are soft edges, not walls: manufacturers didn't switch overnight, exemptions exist, and old stock tags kept getting sewn in. And these are floors: a shirt printed in 2019 carries all three happily. Finding none is the interesting result.

Union labels, and why they beat everything else on a tag

A union label narrows the window from both ends, which almost nothing else on a tag does. That makes it the most valuable thing in a photograph of an American made garment.

The two you meet most are the ILGWU, the International Ladies' Garment Workers' Union, and the ACWA, the Amalgamated Clothing Workers of America. Both were merged out of existence on dates you can check, which is what gives their labels a ceiling.

• No AFL-CIO wording on the label generally puts the garment before roughly 1959 to 1963, because the artwork changed in the years after the AFL and the CIO merged in December 1955. That window is hobby convention, widely repeated across sellers, and no regulation confirms it.
• The union's exact name is the dating tell. ILGWU means before the back half of the 1990s. ACWA, with no Textile in the name, means nearly twenty years earlier still, because the ACWA became the ACTWU in 1976. Read the letters instead of recognizing the shape.

Leftover label stock blurs both ceilings, the way it blurs every edge in this post. Union labels are also small, low contrast and usually photographed badly. If the listing shows the front graphic in four megapixels and the union label as a beige smudge, ask for one flat, in focus photograph of every tag.

Can you look up an RN number to date a garment?

No. An RN number identifies a company, not a year. It gives you a floor and never a ceiling, and reading an RN issue date as a manufacture date is the most common error in the subject.

RN stands for registered identification number. The Federal Trade Commission issues them, and 16 CFR 303.20 lets a company print one on the label in place of the name the rules would otherwise require. Companies apply through the FTC site that section names, rn.ftc.gov. The database you want is the separate public one: search it free at the FTC's RN Database, by company name, legal business name, or the digits alone with the letters dropped.

What comes back is a company. Three things to do with it:

• Take the floor, and only the floor. An RN issued in 1985 means the garment wasn't made in 1975. It could still have been made in 2015, because RNs stay valid and companies keep printing them for decades.
• Check the name. If the tag says one brand and the RN comes back as an unrelated business, ask before you bid.
• Expect a dead end on the oldest pieces. A genuinely early garment can predate the numbering system, so an absent RN is consistent with the rest of the tag and not a reason for suspicion.

The lookup is free and takes thirty seconds. It's also the check almost nobody runs.

Is single stitch proof that a shirt is vintage?

No. Single stitch construction is evidence, it's weak evidence on its own, and reproduction shirts use it deliberately because buyers have been taught to look for it.

The observation behind it is one sellers and collectors report consistently: single needle hems and sleeve cuffs were standard on American blank tees into the 1990s, with double stitching becoming the norm from roughly the middle of that decade. No regulation sits behind that date, which is why it's soft at both ends. As a rough sort it works. As a verdict it doesn't, because anyone printing reproductions can buy blanks made the old way.

So it never stands alone. Combine it with at least two of the following:

• The tag brand on the blank, checked against the era the shirt claims.
• The copyright line under the graphic. Licensed music and film merchandise usually carries a copyright symbol, a year and a licensor name printed on the shirt itself. That line is on the garment and not on a tag, which is why it's harder to fake.
• Fabric weight and hand. Older blanks generally feel heavier and coarser than a modern retail tee.

If those disagree, believe the disagreement. A shirt whose tag says one decade and whose print copyright says another isn't telling you that you found a bargain.

How do you spot a reproduction band tee before you bid?

Run four checks in this order, because the first one catches the expensive fakes.

1. Check the tag against the blank it's sewn into. The expensive reproduction isn't a fake tag. It's a period correct tag cut out of a worthless shirt and sewn into a modern blank. An original is caught in the same seam as the collar binding. A transplanted tag usually sits on its own stitch line, sometimes with the old stitching still visible as holes.
2. Read the print surface. A period screen print sits on top of the fabric with a raised, slightly rubbery texture and cracks in irregular lines that follow how the shirt was folded. A direct to garment print is smooth and flat and sinks into the weave. Ask for one photograph with light raking across the print from the side.
3. Check that the wear is coherent. Age shows up everywhere at once, so collar stretch, hem fraying, shoulder fading and print cracking should agree. A heavily cracked graphic on a crisp, unstretched collar is a mismatch, usually a new print distressed on purpose.
4. Confirm the event happened. Tour tees list dates and cities on the back. Look up the tour. A back print naming a city the band didn't play that year ends the conversation.

Two things are specific to this category:

• Reverse image search still helps and no longer settles anything. It exposes a stock photograph by finding its source, which only works when the picture came from somewhere. Listings built on generated images are documented and increasing, and a generated image has no source.
• The Authenticity Guarantee is unlikely to be your backstop. Its apparel category covers eligible streetwear brands at $200 and over, listed in Men's or Women's Clothing (eBay's Authenticity Guarantee page). A 1980s tour tee sits outside that whatever it sells for. We wrote up the program category by category; for vintage clothing you're on your own.

The order to read an eBay listing in covers the general version of this pass, including the condition field, whose values eBay defines even though the seller is the one picking from them (item conditions).

What the size on the tag is actually worth

The printed size isn't a measurement. It's a marketing decision made by one company in one decade, and fifty years of drift have left it telling you almost nothing about fit.

We're not going to hand you a conversion table. No version holds across decades, brands and countries, and every published one is somebody's average dressed up as a rule. What works instead:

• Ask for flat measurements in inches. On a tee that's pit to pit across the chest, and length from the highest point of the shoulder to the hem.
• Double the pit to pit before comparing it to anything. A 21 inch pit to pit is a 42 inch chest, and mixing the two conventions is the most common way people order a shirt two sizes off.
• Expect the garment to measure smaller than the letter suggests. Sellers and collectors report the drift running one way: vanity sizing moved the letters, so a vintage medium often measures closer to a modern small. That's a reason to ask for numbers, not a conversion you can apply.
• Measure a garment you already own and like. Compare numbers to numbers. It's the only step with no interpretation in it.
• Treat refusal as data. Measuring a shirt takes under a minute and needs a tape.

Vintage is also where price research is hardest. Every piece is effectively one of one: no set number, no sealed box, no graded population to compare against. Our own deal score compares a listing against the asking prices of the other listings the same eBay US search returned, needs at least five of them before it will label anything, and calls something a HOT DEAL at 20 percent or more below their average. It isn't a sold price comparison, because we hold no sold price data, and on a truly unique garment it usually declines to say anything at all. That's correct, and not a fault. Price history has the same shape: it starts the day you start tracking an item, for us and everyone else. How to judge an eBay price without sold data is the longer answer.

A vintage garment is also listed once, by one person, often with the brand spelled the way they say it and not the way it's printed. Searching misspellings is eBay's own suggested tactic, and it works better here than anywhere, because no catalog entry forces the seller to get the name right. You can run that search without an account, and the free tools worth using covers the jobs other people do better than we do.