SIG Sauer Serial & Date-Code Lookup

Enter either marking and the tool answers from its own register: the stamped two-letter German date code (read beside the proof marks, not from the serial), or a US-made serial number, answered only where an observed box-sticker point exists. A serial is never dated by the German ladder and a code is never invented from a serial's shape: SIG publishes no dating chart for either era, and the maker's own serial tools answer recalls, not dates. SIG Sauer is two makers' systems in one name and this page keeps them apart. German-era guns (Kiel / Bundesadler proofs) carry a stamped two-letter date code beside the proof marks — not in the serial: “It's impossible to decode a SIG serial number to determine when it was proofed.” Enter the code and the answer is the year of proof, which the sources warn is not always the manufacture year. US-era guns (SIG Sauer Inc.) carry no code — “Most late German (about late 90's or so) and U.S. made 226's don't have the date code” — and their serial names the model, not the year. For them the only sourced dates are 593 observed box-sticker points, 2005–2023: a serial that matches an observed point gets its date; one that falls between points is told so, with the neighbouring points printed and no invented window. The era is never guessed from the serial's shape, and SIG publishes no dating chart: the maker's public serial tools answer “is my P320 upgraded” and “is my rifle recalled” — a different question than “when was it made”.

How dating a SIG Sauer works — two eras, two registers

Most serial charts give one number per year. SIG Sauer is different, and using the wrong method gives the wrong year — so it is worth knowing what the chart actually asserts.

two registers: German proof date code HG=1976 through AK=2009 (+ printed BC=2012), and US box-sticker points 2005–2023 — the maker's own serial tool checks recalls, not dates

German era — the code. Each letter maps to one digit through the printed alphabet A=0 B=1 C=2 D=3 E=4 F=5 G=6 H=7 J=8 K=9 — SIG prints J for 8 and never I (“Sig did not use ‘I’ in date codes as it is too easily confused with the number 1”). First letter = decade, second = units. The ladder printed by the sources runs HH=1977 through AK=2009, with the per-gun rows of the sigforum classic list supplying HG=1976 and a photo caption printing the single post-2009 pair BC=2012; the table's own cell after AK reads UNKNOWN, so BA, BB, BD and every later pair is refused by name, not extrapolated. The same honesty applies before 1976: HA–HF fit the structure but no opened source prints them, so they refuse too. A pair containing I is refused: the NRA's generic German chart (I=8, no J) contradicts the SIG-specific trio, that contradiction is kept on this page, not averaged, and the NRA reading is never borrowed onto a SIG. The code dates the year of proof at the proof house: “Some P230/2 SIG's were proofed at a different time then when manufactured.”

US era — the points. The sigtalk 54176 thread collects owners' box stickers: masked serial; sticker date; model; contributor. Every count on this page (dated points, model designations, the silent rows) is computed live from the register as delivered, updated January 24, 2023 (printed), opened 2026-10-20. This is a collector tier register, stated as such: “not available elsewhere” says the thread itself, and the maker publishes none of it. The thread's own data is the proof that a SIG prefix is not a date: masked prefixes recur across different sticker years (the count and a spanning example are printed below), and prefixes like 37a → P220, 47a → P226 name models. So the register answers at observed points only: a serial matching a printed point gets that observation and its credit line; a serial between points gets the neighbouring points and the word between; nothing in between the two is ever printed as a window. The thread's one rejected line (EAK067xxx, 2001 — outside its 2005+ scope) is recorded, not silently dropped.

The guard. The two registers answer different markings from different companies (the two firms have been organizationally separate since 2000 — Wikipedia, opened 2026-10-20). A marking whose era is ambiguous — a serial-shaped input whose first two letters are themselves a printed code pair, like KD1234 — is refused, never routed by guessing. The database's country column is never an era selector either: 56 rows say “Germany” and carry a year ≥ 2005 (measured 2026-10-20).

The maker's own serial tools were opened 2026-10-20 and quoted verbatim — and they are the negative. SIG SAUER publishes a P320 upgrade checker (“Enter your serial number to the right to see if your P320 has already been upgraded.”), a rifle safety recall entry (“This serial number IS affected by the recall.” / “IS NOT affected”) and a CROSS recall list. None of them answers “when”. The maker holds the date and does not publish it: an owner who phoned customer service was told “manufactured in Jan. 2013” over the phone (sigtalk 50131, opened 2026-10-20) — internally yes, publicly no. On the points side the worked examples are the observations themselves: every point on this page carries its printed line and its credited contributor. On the code side the corroboration is a field reading: an owner's 1993 grip date and “The KD stamp on the slide confirms this” (sigtalk 498, opened 2026-10-20) — the decoder answers KD as 1993.

593 observed US box-sticker dates (US-made / US-assembled production, 2005 onward per the thread's own stated scope) — the second register

One owner's box sticker: masked serial (last three digits X'd by design), production date, model, contributor. The date type is what the sticker carries; the maker publishes none of this.

A query that falls between documented points gets the neighbouring points and the statement that it falls between them. NO interpolation, NO implied range.

And the maker's own serial tools? The maker DOES publish serial-number tools, they are public, and they answer a DIFFERENT question — recall/upgrade status, never date of manufacture. This is the honest first answer about a SIG serial.

• P320 voluntary upgrade checker — it answers: has this P320 already been upgraded (drop-safety programme) — not when was it made. Printed on its page: “Enter your serial number to the right to see if your P320 has already been upgraded.”
• Rifle safety recall entry — it answers: is this rifle affected by the two-stage-trigger recall — not when was it made. Printed on its page: “Enter your firearm’s 9-digit serial number in the box below.”
• CROSS recall list — it answers: is this CROSS rifle on the recall list — not when was it made. Printed on its page: “Please enter your serial number. If your serial number is included in the recall you will be directed to a product return registration page.”

Contradictions kept, not averaged:

• alphabet_I_vs_J — SIG proof codes use J=8 and never print I: alphabet A B C D E F G H J K (carolina_sticky + sigforum_classic_list + sigtalk_p230_ladder, all opened 2026-10-20); explicit note: 'Sig did not use “I” in date codes as it is too easily confused with the number 1'. Against: The generic GERMAN MODERN DATE CODES chart prints A B C D E F G H I K -> 0..9 (I=8; J absent) — nra_serialization_pdf, GERMAN MODERN DATE CODES page, text-extracted 2026-10-20 ('IF 85 or 1985'). Shipped resolution: SIG-specific sources win for the SIG map (J=8, no I) because three SIG-specific sources agree and one forum member frames the national split ('SIGs exclude the letter I, where HKs exclude the letter J', sigtalk_498). The NRA chart governs German proof codes GENERALLY (H&K/Walther class) and is kept here un-merged. A code pair containing I on a SIG is REFUSED (pair_outside_alphabet) rather than silently read via either alphabet.
• prefix_58a_model — '58a - P290' — the thread's own prefix table (sigtalk_54176, opened 2026-10-20). Against: 'With some exceptions, most U.S. Domestic Consumer P320 serial numbers begin with “58A”, “58B”, or “58C”' — the MAKER (maker_p320_upgrade, opened 2026-10-20). Shipped resolution: Neither is dropped. The maker statement is maker-tier, the thread is collector-tier; the file ships both verbatim and the engine must NOT use 58A to name a model without a warning. What the prefix is not: a date (measured prefix_is_not_date_evidence).
• year_100_block_location — Batch body: '136 rows ... names of the form AA P220, AA P221, ..., AA P270 (id: aa-p220 ...)'. Against: Re-measurement today (db_measurement.fabricated_year100): the year-100 block runs AA P263..AA P400 (ids aa-p263..aa-p400 with the printed gaps); AA P220..AA P262 are the 41 AA-prefixed rows that carry real-looking years (38 of them 1980). Shipped resolution: count identical (136), location corrected with the id list. The prefix alone is not the defect; the impossible year is — as the body warned.

If your model is not listed, SIG Sauer has not published a serial chart for it — the tool will say so rather than guess.

Questions

Why does my serial number not give a year?

It falls in a gap between the published ranges, or in a model year the chart does not cover — the tool will show the nearest documented range instead of guessing. A public lookup should not invent an answer. For a SIG Sauer: the German gun may carry no code at all — “Most late German (about late 90's or so) and U.S. made 226's don't have the date code” is a documented normal outcome, not a failed read; the pair may contain I (the SIG alphabet never prints it — refused, the NRA chart is not borrowed); it may sit in the unprinted HA–HF stretch or anywhere after AK=2009 except BC (refused, the gap is named, the ladder's UNKNOWN cell is shown on this page); a US serial may match no observed point (refused with the neighbouring points, never interpolated); its prefix may carry observations on different years (refused — the thread's own G508XXX spans 2010 to 2013); the sticker may no longer exist (its survival is not established for any given gun); or the input may straddle both systems (refused as era-ambiguous). None of these is rounded to a neighbouring year, and no pre-1945 “SIG” is dated by either register — the brand partnership starts 1975 and the Swiss lineage is a different system.

How accurate is this?

The ranges are reproduced exactly as published in SIG Sauer's own chart. That is not a factory letter — for a definitive date, the factory records are the authority. The code ladder is a collector compilation re-opened 2026-10-20 (the sigtalk P230/P232 ladder, updated by its OP 03/07/2026 per its own print; the carolinaShooters sticky; the sigforum classic list, which today sits behind a login wall and was read through its 2026-06-08 Wayback snapshot; realgunreviews, 2016). Four independent sources, zero maker prints: that is an honest grade, not a factory record — and the code dates PROOF, which the compilation warns may differ from manufacture. The points register is 593 owner observations 2005–2023 (thread updated January 24, 2023 — printed): a crowd-sourced record of box stickers, tier stated, one line rejected out of scope and recorded. Of the 354 SIG rows in the database here, 218 are worth dating (measured 2026-10-20); 136 carry an impossible year of 100 and are excluded from coverage, not silently counted — see the coverage note on this page. Nothing either register prints is filled by guesswork.

Where does the data come from?

SIG Sauer's official serial-number reference — nothing else. Every chart this tool uses is published, and every page names the publisher it came from — the maker where the maker publishes one, a named association or author where it does not.

Other makes

Serial-number charts are published for:

• Browning serial number lookup
• High Standard serial number lookup
• Ithaca serial number lookup
• Purdey serial number lookup
• Remington serial number lookup
• Ruger serial number lookup
• Savage serial number lookup
• Smith & Wesson serial number lookup
• Winchester serial number lookup