Perfume Batch Code Checker: What the Numbers Mean

A perfume batch code checker reads the production year from the code on the bottle, not whether the perfume is real. Type the code to decode its shape and year.

How Perfume Batch Codes Work

A batch code is a short stamp that identifies one production run of a fragrance. It is not a serial number. Every bottle filled in the same run carries the same code. The code appears twice: once on the base of the bottle and once on the bottom of the box. The two must match. The code is required by law in Europe but its meaning is not, which is why every house writes it differently.

The legal duty comes from Article 19 of Regulation (EC) 1223/2009, the EU cosmetics regulation. It lists “the batch number of manufacture or the reference for identifying the cosmetic product” among the items required on the container and the packaging. The same article explains why perfume rarely carries an expiry date. A product with a shelf life over 30 months need not show one. It shows the open-jar symbol with a month count instead. In the United States the FDA’s summary of cosmetics labeling requirements lists product name, net quantity, firm name, country of origin, ingredients and warnings. A batch number is not among them. Perfume sold in the US carries one anyway because the same bottles ship to Europe.

No fragrance house has published how its code is built. Everything a batch code checker knows comes from collectors comparing thousands of codes against purchase receipts, box designs and regulatory changes. That reverse engineering has produced reliable readings for a handful of manufacturing groups and guesswork for the rest. The checker on this page separates the two. It names the group whose shape your code fits. It reads the year where two independent sources agree on how, and says so when they do not.

Fragrance houses fill their bottles in a smaller number of factories than the brand names suggest. Estée Lauder Companies fills Tom Ford, Jo Malone London, Le Labo and Kilian Paris, and all of them share one three-character code. L’Oréal fills Lancôme, Yves Saint Laurent, Giorgio Armani, Prada and Maison Margiela, and all of them share one six-or-seven-character code. LVMH’s perfume arm stamps Dior, Guerlain and Givenchy the same way. Knowing the group is the first step in reading the code, because the group, not the brand on the bottle, decides the format.

Batch Code Format Table by Manufacturing Group

Use the table to place a code before reading anything else. Positions are counted left to right as the code is printed. Confidence records how many independent sources agree on the reading.

Three shapes overlap, and the checker reports both readings when they do. A four-digit code is a counter on Chanel and a year-first code on Coty. A five-character code of two digits, a letter and two digits is Creed from 2010 to 2022 and Puig from 1990 to 2000. Both read the first two digits as the year, so the overlap is harmless. A digit followed by a letter and two digits is LVMH and Puig, and both read the first digit as the year, though only LVMH’s month letter has been decoded.

The letter tables in the LVMH and L’Oréal rows are where public decoders disagree most. For Dior, Raiders of the Lost Scent’s guide gives A to M with I skipped for 1998 to 2007 and again from 2018, and N to Z with O skipped for 2008 to 2017. A second decoder gives A to L with no skip. The two agree on the year digit and on the month for January to August, and differ by one month from September on. The checker reports the year, which both agree on, and gives the month from the Raiders table because it is the only one built from dated bottles across three houses.

Reading the Three Shapes You Will Meet Most

Most perfume on the second-hand and discount market comes from Estée Lauder, LVMH or L’Oréal factories, so these three readings cover most codes typed into a batch code checker. Take one worked example for each.

1. Estée Lauder Companies: A63

1. Ignore the first character. A is a product line or filling site. It never carries a date.
2. Read the second character as the month. 1 to 9 are January to September. A is October, B is November, C is December. Here 6 is June.
3. Read the third character as the last digit of the year. 3 is 2023 or 2013.
4. Use the launch date to pick the decade. A Tom Ford Private Blend launched in 2019 cannot carry a 2013 code, so A63 is June 2023.

Both Fragnatium’s Tom Ford page and the COSR guide to Estée Lauder codes give this reading. COSR lists dated examples such as c79 for July 2019 and ac3 for December 2023. The month character is the part that catches fakes: a code with 0 or D in the second position is not an Estée Lauder code, whatever the bottle says.

2. LVMH: 3E01

1. Read the first digit as the last digit of the year. 3 is 2003, 2013 or 2023.
2. Find which half of the alphabet the letter belongs to. A to M was used from 1998 to 2007 and again from 2018. N to Z was used from 2008 to 2017. E is in the first half.
3. Combine the two. A first-half letter rules out 2013, so 3E01 is 2003 or 2023. The bottle’s design settles which.
4. Read the letter as the month within its half. In the first half A is January and M is December with I skipped, so E is May. In the second half N is January and Z is December with O skipped.
5. Treat the last two digits as a run number. 01 is the first run of that month.

Raiders gives 8A01 as January 1998, 8P01 as February 2008 and 0T02 as August 2010. Its Guerlain and Givenchy guides use the same halves, with 3B01 for 2003 and 8R01 for 2008. The I and O skips matter for authentication: a Dior code containing either letter has been misread or misprinted.

3. L’Oréal: 38U60OG

1. Read the first two digits as a plant or line. 38, 40 and 62 are the values seen most. They carry no date.
2. Read the third character as the year. The letters run H 2011, J 2012, K 2013, L 2014, M 2015, N 2016, P 2017 and R 2018. They continue S 2019, T 2020, U 2021, V 2022, W 2023, X 2024 and Y 2025. I, O and Q are skipped. U is 2021.
3. Stop there. The remaining characters have not been decoded and do not give a month.

The letter table is assembled from two sources that do not overlap in the years they cover and do not conflict. Raiders documents H for 2011 through R for 2018 on Yves Saint Laurent bottles. Fragnatium’s YSL page documents S for 2019 on YSL and Y for 2025 on Prada. The five letters between S and Y fill 2020 to 2024 in order, with no skip left to place. Fragnatium withdrew its month reading after finding that the only source claiming one published formats that did not fit real bottles, so the checker does not give a month either.

Where To Find the Batch Code on a Perfume

The batch code is printed in two places, and finding both is part of the check. On the bottle it sits on the base, either ink-jetted in small dark characters, laser-etched into the glass, or printed on a small sticker. On the box it is printed on the bottom flap or a lower side panel, usually near the barcode. It is in a different typeface and shorter than the barcode number.

The bottle code is the harder one to read. Ink-jet codes on clear glass show best against a white sheet of paper held behind the base. Etched codes show best under a raking light from a torch held at a low angle. A 10x loupe makes a four-character Dior code legible; a phone camera at full zoom does the same job.

Several things printed near the code are not the code. The 13-digit number under the barcode is the EAN and is identical on every bottle of that size. A 13-digit string typed into a batch code checker will be rejected as a barcode. The concentration and volume, for example “EDP 100 ml”, are product facts. A retailer’s sticker with a SKU or a price is not the code either. The batch code is the shortest alphanumeric string on the base, and it is the only one that also appears on the box.

Travel sprays and discovery sets are stamped differently. A 10 ml atomiser usually carries the code on the metal collar or on the sleeve rather than the base. Some decant vials carry no code at all, because the reseller filled them from a larger bottle. A decant with no code is normal; a full retail bottle with no code is not.

Batch Code vs Barcode vs Reference Number

Three numbers sit on a perfume box, and only one of them dates the bottle. The table separates them so the right one goes into the checker.

The barcode is the most common wrong entry. It is long, it is on every box and it looks official, but it says nothing about when a bottle was filled. Two bottles with the same barcode can be twenty years apart. The reference number is the second most common wrong entry, and on some houses it looks like a batch code. That is why the checker asks for the string that appears on both the bottle and the box.

The rule that catches most fakes is not in any table. The batch code on the bottle and the batch code on the box must be the same string. Genuine bottles are boxed on the same line that fills them, so the two codes are printed from the same run. A counterfeiter buys or copies one genuine box and pairs it with bottles stamped from a different template, and the codes drift apart. A mismatch does not always mean a fake, because a retailer may have re-boxed a tester, but it always means the box cannot vouch for the bottle.

Why Batch Code Checkers Disagree With Each Other

Public batch code checkers are databases of reverse-engineered formats, and where the reverse engineering is thin they fill the gap with a rule that looks confident. Comparing nine of them against collector-documented bottles for this page turned up three kinds of error that a reader should know about.

The first is a format that fits no real bottle. One widely ranked guide describes L’Oréal codes as five characters, a year letter followed by a three-digit day and a year digit, with A for 2021. Real L’Oréal codes such as 38U60OG, 40S51Z and 62S113 are six or seven characters with the year letter third, and U is 2021 on dated bottles. A checker built on the five-character rule will call every genuine Lancôme code invalid.

The second is a table that contradicts documented batches. One decoder gives Creed a leading year letter, A for 2010 through P for 2023, and decodes M5K23 as 2021. The Aventus batches that collectors have tracked for a decade, 11Z01, 14T01 and 17Q11, all begin with two year digits, as ScentLedger’s batch guide sets out. Under the letter-first rule 14T01 would be an invalid code.

The third is a wrong manufacturer. The same decoder states that Versace fragrance has been made by Coty since 2003. Euroitalia’s own company history records the acquisition of the Versace licence in 2005, alongside Moschino from 1988 and Michael Kors from 2022. A checker that assigns Versace to Coty will apply Coty’s four-digit rule to a seven-digit Euroitalia code and return a date for a shape it does not understand.

The honest checkers say what they do not know. Fragnatium lists 17 houses, including Chanel, Dior, Creed and Parfums de Marly, for which it offers guidance only. Its reason is that any date would come from community-built tables “which disagree with each other and lose accuracy every time a brand quietly revises its stamping”. SkinCarisma states that it has never been able to read a L’Oréal code. That candour is a better sign of a reliable tool than a confident date.

For Chanel the disagreement is structural rather than a mistake. Three independent accounts, the ex-employee letter published by Raiders, ScentX’s 2026 Chanel guide and Fragnatium’s Chanel page, agree that the modern four-digit code is a counter with no year inside it. Any checker that returns a year for a Chanel code is reading a lookup table someone built from purchase dates. Fragnatium notes that Chanel client care will confirm a production period from the code on request. That is the only authoritative route.

What This Checker Cannot Tell You

A valid batch code does not mean a genuine perfume, and this is the most important limit of any batch code checker. Counterfeiters copy codes from real bottles and print them on fakes. A copied 3E01 passes every format test because it was a real code to begin with. Format checking rules out badly made fakes with impossible codes, such as a Dior code with an O in it or an Estée Lauder code with a zero month. It does not rule in a genuine bottle.

The scale of the copying is documented at the border. U.S. Customs and Border Protection reported a single seizure of more than 8,500 counterfeit designer perfumes in Fort Lauderdale, with a retail value over one million dollars. An earlier Long Beach seizure held 3,738 bottles of counterfeit Dior and Chanel shipped from Hong Kong in a container declared as phone accessories. The bottles that clear the border carry plausible codes. In February 2026 Estée Lauder Companies sued Walmart in federal court in California over marketplace listings it said were counterfeit Le Labo, Tom Ford, Clinique and La Mer. The company had bought and tested the items first. Estée Lauder’s counterfeit guidance is to buy only from the authorised retailers listed on each brand’s site.

This checker cannot tie a code to a bottle. No house exposes its production records. Creed’s FAQ asks customers who suspect a fake to send photographs and “the LOT number (found at the bottom of the bottle or box)” to customer service. That is the only route to a house-level answer for that brand. Chanel client care does the same for Chanel. For every other house, the batch code identifies a run and nothing more.

This checker cannot smell the juice or weigh the bottle. Colour, the fill height in a sealed bottle, the weight of the cap, the sprayer’s throw and the cellophane seam are physical checks. They decide most real cases. A perfect code on a bottle with a plastic dip tube where the house uses glass is still wrong. Our fragrance guides cover those checks house by house.

This checker cannot date Chanel and cannot read Creed codes stamped after 2023. It gives a probable rather than proven year for Puig, Coty, Euroitalia, Hermès, Mancera and Xerjoff, where the reading rests on a single collector source. Where the checker says caution, the year is a hypothesis to test against the box, not a fact.

Finally, this checker stores nothing. The code is matched in your browser against the patterns described on this page and is not sent anywhere.

From Batch Code to Physical Check

The batch code is the first filter, not the last. A sensible order for a second-hand or discount purchase runs in five steps.

1. Find both codes. Base of the bottle, bottom of the box. Photograph both.
2. Confirm they match. Character for character. A mismatch ends the check.
3. Decode the shape here. Confirm the code fits the house’s documented shape and that the year is possible for the fragrance and the bottle design.
4. Check the packaging against the year. The open-jar symbol has been mandatory in Europe since 2005. Allergen lists arrived between 2003 and 2005 depending on the house. Box addresses change with corporate moves. A 2023 code on a box without an allergen list is wrong.
5. Get a physical opinion for expensive bottles. A reseller with a money-back authenticity guarantee, or the house itself for Creed and Chanel, should see the bottle.

The tools index lists the other code lookups on this site. The physical side of a bottle, meaning the cellophane seam, the cap weight, the sprayer geometry and the fill height, is in fake perfume: how to tell, and a code that reads old rather than fake usually means grey-market stock, which the FragranceNet review explains in detail. The same copied-code problem appears in watches and bags, where the Rolex serial number lookup and the Louis Vuitton date code check both read format and era only. The Coach serial number lookup faces the same shared-code limit as a perfume batch, because every bag of one style carries the same number.