Taurus Serial Number Date Lookup

Enter a Taurus serial number. Serials that begin with the post-October-2019 letter pair get a plain answer: AA = 2019, AB = 2020 … AH = 2026, third letter A-L = month — printed by one collector source reporting a manual update, with per-pair provenance shipped with the answer. Every older serial gets the other honest product this page sells: the documented refusal — the two surviving readings quoted side by side, the maker's own retired lookup with its death dates, and no guessing. Taurus publishes no serial chart today; the numbers in the refusals are measured, the quotes are verbatim. Taurus is one printed rule plus a pile of contested folklore, and this page keeps them apart. From October 2019 the serial begins with a letter pair that encodes the year (AA=2019 … AH=2026) and an optional month letter A-L — stated by Erick Tamberg, answered directly by Dr. Domingos Tocchetto, author of the official Brazilian Taurus/Rossi manuals (“ACD means April, 2021”), and corroborated by owners’ dated purchases. That ladder is printed AA-AD; AE-AG are projections from the printed rule (labelled on every answer) and AH is observed in the wild. Everything older refuses: the 1981-2006 and 2007-2019 prefixes are read year-first by one camp and caliber-first by the other, each camp’s worked examples fail the other (SDN/TDM vs HI/KI), and the maker’s own retired tool agreed with none of them (NFX → “Feb 2013”, a date no chart prints). Pre-1981 numeric serials: not established — the Miami records were microfiched in the 2019 move. The tool never averages the two readings and never silently resolves I-as-1 or O-as-0.

What this register covers — measured. Coverage, measured 2026-10-21 in this working copy: of 265 rows labelled Taurus in the database here, 265/265 are 'Brazil', 202 pistols and 63 revolvers; 143 rows carry a year field, span 1980-2043. 16 carry impossible years (2027, 2028, 2029, 2030, 2031, 2032, 2033, 2034, 2035, 2036, 2037, 2038, 2039, 2040, 2042, 2043), enumerated in the source measurement: taurus-pt-129=2043, taurus-pt-128=2042, tarasus-pt-126=2040, taurus-pt-125=2039, taurus-pt-124=2038, taurus-pt-123=2037, taurus-pt-122=2036, taurus-pt-121=2035, taurus-pt-120=2034, taurus-pt-119=2033, taurus-pt-118=2032, taurus-pt-117=2031, taurus-pt-116=2030, taurus-pt-115=2029, taurus-pt-114=2028, taurus-pt-113=2027. The tool answers the serial's printed prefix, not these rows: no DB year was used as evidence for the ladder.
Look up your Taurus serial number

Enter the serial exactly as stamped. 2019+ guns whose serial starts AA-AH get a year (projected pairs labelled); everything else gets a named refusal — both outcomes are sourced, neither is a guess.

Open the Taurus serial lookup →

How dating a Taurus by serial number works

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

The one printed rule: the post-2019 A-prefix year pair (AA = 2019 to AH = 2026, third letter A-L = month)

The rule. year = 2018 + position_of_second_letter(A=1). AA=2019 ... AH=2026. First letter fixed A; second letter = year; optional third letter = month: A=January B=February C=March D=April E=May F=June G=July H=August I=September J=October K=November L=December. Each answer carries its provenance status straight from the register: printed (AA-AD, the pairs the sources print), observed (AH — the AHG revolvers; the claim’s own thread predates July 2026 and the hub prints that fact), projected (AE-AG — arithmetic of the printed ladder, no source prints them; the register makes labelling mandatory, not optional). What happens when the second letter exhausts Z (projected 2044) is NOT established: AI and beyond refuse.

What the month letter is worth. The printed month letter disagrees with the gun’s own stamps on real guns — ACL (serial says December, frame and box say October 2021), ABJ (says October 2020, bought September 2020), AHG (“July 2026” reported in August 2025). A month answer therefore ships as “the month letter says X” with the conflict stated, never as a bare fact.

Where the letters come from. The ladder is collector-tier and stated as such: one collector source (Erick Tamberg) reporting the update of Dr. Domingos Tocchetto’s manual — Tocchetto authored the maker’s own Brazilian manuals, but reaches this record through Tamberg’s posts — plus specimens with independent purchase dates (ABD, ACL, ABJ), and zero maker-published rows: Taurus prints no chart today (verifiable negative, measured 2026-10-21). The claimed carrier “Seventh Edition of the official Taurus manual” is recorded as relayed by one poster in two threads, not verified against the book.

AA = 2019 — printed
AB = 2020 — printed
AC = 2021 — printed
AD = 2022 — printed
AH = 2026 — observed (single, self-contradicting claim)

The projection band — no source prints these pairs. These 3 rows follow from the printed ladder rule alone; the tool answers them with this label (choosing to label rather than silently extend, as the forums did, or silently refuse). The register's own line on these pairs, quoted verbatim: “Printed as pairs: AA AB AC AD (2019-2022). The only wild report beyond them is AHG (AH = 2026) - a single, self-contradicting claim, printed with its oddity rather than as fact. AE-AG are projections from the printed ladder rule and are labelled 'projected' wherever they answer. What happens when the second letter exhausts Z (projected 2044) is NOT established - pairs beyond AH reach a named refusal instead of an extrapolation.”

AE = 2023 — projected
AF = 2024 — projected
AG = 2025 — projected

The worked examples are the sources’ own. ADJ → “AD means 2022… J means October” (taurusarmed 443010, opened 2026-10-21); ACD → “ACD means April, 2021”, posted by Erick Tamberg as Dr. Tocchetto’s answer (438919); ABD → April 2020, corroborated by a June 2020 purchase (441192). The decoder answers all three exactly. And the maker’s own tool is the negative: find-model.cfm printed “Enter the serial number … to identify your model” — a per-serial database, never a rule (Wayback: 104 captures with HTTP 200, last 2018-12-12; first archived 404 2021-04-21; live 404 measured 2026-10-21). Its successor /support/serial-number-lookup/ archived 200s only 2019-04-01 to 2020-11-26. NFX got “built in Feb of 2013” from that tool while every chart demanded NGN — proof the tool implemented a database, not the folklore. What survives is not a replacement: the live /support/repair-status/ serial search proxies SAP repair records — SerialID, ProductID, status — never a date. It is not proxied here and not described as a dating service.

The register: 2 published readings, 8 ladder pairs (4 printed, 1 observed, 3 projected), 8 named refusal classes

Everything below is quoted from the document this decoder is pinned to (sha256-checked on every request). Posters and dates are the sources’ own; nothing here is paraphrased for comfort, and nothing is averaged across readings.

Reading 1 - first letter is the YEAR (two-letter era scheme) — scope as claimed: serial numbers preceded by letters, 1981 onward.

Printed rule, verbatim: “FIRST letter means the YEAR, beginning with A in 1981 to Z in 2006. Returned to A in 2007, changing the second letter for month. SECOND letter means the MONTH. A to L between 1981-2006. M to X from 2007.” — Erick Tamberg, 'Tips for Dating Taurus Revolvers' post #1 (2021-08-05), quoted verbatim by JMD669 in thread 443010

Worked examples that check:

  • HI prefix = 'September 1988' (+P blessing, JMD669 2024-04-20): H is the 8th letter -> 1981+7=1988; I is the 9th letter -> month 9. Arithmetic checks.
  • KI = 'K = 1991 I = September' (75526 page-3 answer): K 11th -> 1991; I 9th -> September. Arithmetic checks - but the serial behind it was printed 'K179xxx' and the answerer silently read 1 as I.
  • A continuation table in thread 103331 (2013) prints the same year ladder A-1981...Z-2006, then A-2007...M-2019...Z-2032 - which COLLIDES with the 2019 scheme (AA=2019).

Worked examples this reading cannot explain:

  • TLA3385S PT92 'purchased new in 1984': reading T as a year letter gives 2000 (or 1981+19=2000), not 1984 (75526).
  • Any 3-letter pistol serial from the 2007+ era where the community reads the FIRST letter as caliber (see Reading 2) - the same letter position cannot be both year and caliber.
Reading 2 - first letter is the CALIBER, second is the YEAR, third is the MONTH — scope as claimed: three-letter pistol serials from ~2007 ('applies only to the Mil-Pro line', later generalized).

Printed rule, verbatim: “First Letter = Caliber T = 9mm N = .45 S = .40. Second Letter = Year A = 2007 B = 2008 C = 2009. Third Letter = Month O = January P = February Q = March R = April S = May T = June U = July V = August W = September X = October Y = November Z = December.” — MTee, thread 'Taurus Date of Manufacture or Shipping????' post 2009-07-28 (#10)

Worked examples that check:

  • SDN: S=.40, D=2010 (A=2007 ladder), month by M-January table = February -> matches the maker phone answer 'true date of manufacture was February, 2010'. (The same gun's website answer at the time was 'production year 2011'.)
  • TDM26745: T=9mm D=2010 M=January (poster accepted, 2024-04-20).
  • TGM: T=9mm G=2013 M=January (owner claim on bubbleloading).

Worked examples this reading cannot explain:

  • NFX: maker tool answered 'built in Feb of 2013'; under this chart the poster expected NGN, never NFX (103331, 2013-09-04).
  • TJY53XXX: 'so the last month chart doesnt work' (JJFII, 75526 page 3, 2018-10-13).
  • TYE: bought early 2007, second letter Y fits no printed year ladder, third letter E is not a month letter in any O-Z table (22412, 2009-07-28).
  • TZG claimed '9mm, 2006': 2006 is Reading 1's Z-year value, the T=9mm is Reading 2's first letter, July (G) is the A-L month table - one post, three schemes glued together (75526).

And its own month tables disagree:

  • O_is_January: MIL-Pro chart, 22412 (2009-07-28): 'O = Janurary ... Z = December', and the owner self-checks 'my Mil-Pro is a 9mm (T) and was manufactured in 2008 (B) during the month of October (X)'.
  • M_is_January: SDN512XX owner (75526, 2013-02-09) reports phone answer February 2010 for SDN (M=January scheme), and JMD669 reads TDM as 'T=9mm D=2010 M=January' (2024-04-20); bubbleloading's owner reads TGM as Jan 2013; 438919 prints 'M=January (skip Q and V letters) if after 2007'.

The 8 refusal classes — what this tool answers when it does not answer (each is a documented position, not a shrug):

  • serial has no letters at all — Numbers-only serials predate the alphanumerical system (J-frame revolvers from 1981; K-frame only from November 1988; pistols vary). The Miami-era records were put on microfiche during the 2019 move and the maker's own replies say its records start at 1981 or cannot help. No published source dates a numeric serial. Nothing is inferred. (not_established)
  • two-letter prefix starting B..Z (JB, OF, RA, BO, KI ...) — Two-letter prefixes are contested: Reading 1 (first letter = year, A=1981..Z=2006, restart A=2007) versus Reading 2 (first letter = caliber). One reading camp answers this prefix one way, the other another way, and the community's own answers mix the two inside a single post. This lookup will not pick a side: the sources and the arithmetic for both readings are printed on this page.
  • three-letter prefix starting B..Z (TBS, SDN, TJY, TYE, NFX ...) — For 2007-2019 pistol prefixes the two charts disagree about what the FIRST letter even means (year vs caliber), the caliber camp disagrees internally about the month table (O=January vs M=January vs M-X skipping Q,V), and the maker's own retired tool agreed with none of them (NFX -> 'Feb 2013'). Refused with both readings named.
  • three-letter prefix whose SECOND and THIRD letters fit no published ladder or month table under either reading - the printed cases TYE and TJY — Your prefix contains a letter that no published table explains. TYE is the case the forums reported (thread 22412, 2009-07-28: “SN starts with TYE then 86xxx. I bought it in early 2007 so maybe ‘Y’ as a second digit is 2006. But then the ‘E’ in the third position doesn’t fit … maybe 2006 and earlier is a whole new ballgame”): under the caliber-first chart the second letter has no year value at all (that ladder runs A=2007 upward), and under the year-first reading Y would mean 2006 - while the third letter E is not a month letter in any of the printed O-to-Z or M-to-X tables. TJY is the same shape, and its owner reported the tables failing it too (“well my Serial number is TJY53XXX, so the last month chart doesnt work”, thread 75526 page 3). Neither published reading explains your serial, so whatever the prefix means it sits outside what the charts print - the year is refused rather than fitted by force.
  • prefix contains I/1 (or O/0 where letter-vs-digit changes the reading): KI vs K1, TIE vs T1E, BO vs B0 — On Taurus frames the letter I and the digit 1 (and O and 0) are near-indistinguishable - owners report needing a magnifying glass, and even the forum expert silently resolved K179xxx as 'KI'. If the reading depends on which character it is, we refuse instead of guessing.
  • prefix matches 1 + two letters (1FO, 1GA, 1GC) — Digit-prefixed 1-series serials have no printed decoding rule. The phone line answers them per serial (1GC -> 'April 2022, Brazil'), grip stamps disagree (1GA on a 03/20 dust cover), and the folk etymology ('1GA = first model made in Georgia') is contradicted by 1FO serials from 2017. Refused as not_established.
  • A-prefix with second letter I..Z (AI onwards) — The printed ladder reaches AH=2026; AE-AG are projections and AI+ are extrapolations into an unknown rollover (what happens after Z is not documented anywhere). The engine answers AA-AH and refuses the rest as beyond the printed span.
  • anything only the per-serial database could answer (pre-2007 pistols, 1-series, disputed serials, service history) — Only the maker's own per-serial lookup could ever answer these. It is retired (find-model.cfm 404; /support/serial-number-lookup/ 404; verified 2026-10-21), the Miami-era records are microfiche-only, and what survives at /support/repair-status/ answers REPAIR RECORDS, not dates. The tool is gone - this site says so plainly rather than proxying a 404 or relinking it as live.

Contradictions kept, not averaged:

  • letter_position — On the same 3-letter pistol prefix the FIRST letter is read as YEAR (Reading 1) or CALIBER (Reading 2). Both camps publish worked examples; neither reading explains the other's examples (SDN/TDM favour Reading 2; HI/KI favour Reading 1). They cannot both hold for one string.
  • month_tables_internal — Reading 2 carries three incompatible month tables: O=January (MIL-Pro 2009), M=January (SDN/TDM/TGM lineage), 'M=January skip Q and V' (th_627). The same scheme disagrees with itself about its own month letters.
  • restart_vs_continuation — Reading 1 itself exists in two variants: 'returned to A in 2007' (th_tips) versus the continuation table 'A-2007...L-2018, M-2019...Z-2032' (th_103331). They collide from 2019 on - and the 2019 three-letter scheme (AA=2019) refutes the continuation variant.
  • charts_vs_maker — The retired maker tool's per-serial answers agree with NO chart: NFX -> 'Feb 2013' (chart expected NGN); SDN -> website said 2011 while the phone said February 2010; 'on all 4 of my weapons it is incorrect per Taurus website'. The charts are folklore reverse-engineered around an opaque database.
  • months_vs_stamps — Post-2019 serial month letter disagrees with the gun's own box/frame stamps on real guns: ACL serial says December, frame and box say October 2021; ABJ says October 2020, purchase September 2020; AHG yields July 2026 from a post dated August 2025.
  • maker_page_wording — The retired maker tool's own printed promise was only 'identify your model' - never 'give you a year'. Owners got years anyway ('year of birth; 1992'). A lookup page that promises 'model finder answered dates' overstates what the maker printed.

The maker’s own serial tools, dated:

  • find-model.cfm 'Find My Model / Model Number Lookup System' — retired; first seen 2009-06-16 (Wayback), last archived HTTP 200 2018-12-12, first archived 404 2021-04-21; today: HTTP 404 (2026-10-21)
  • /support/serial-number-lookup/ — retired; first seen 2019-04-01 (Wayback), last archived HTTP 200 2020-11-26, first archived 404 2022-03-17; today: HTTP 404 (2026-10-21)
  • /support/repair-status/ serial search (SAP ServiceRequestCollection) — live — answers: repair-record status for serviced serials - SerialID/ProductID/Status, NEVER a manufacturing date (endpoint probed 2026-10-21)
  • taurusexport.com FAQ link to the retired lookup — live page, dangling link (2026-10-21)
What the maker holds today. Both URLs return HTTP 404 (verified 2026-10-21). The maker publishes NO serial-number dating service and NO printed date chart today: 'serial' does not appear on the homepage, /support/ or /support/faq/ (catch-all verified). The only live serial feature (/support/repair-status/ serial search, plugin sap-serial-search) proxies SAP C4C CRM ServiceRequestCollection and answers SerialID/ProductID/ProductDescription/repair-status - a repair record, not a date, and only for serials that have been serviced. The maker's international FAQ (taurusexport.com, HTTP 200 today) still links the retired lookup page - a dangling link printed by the maker itself.

Why owners have no answer from the maker: When Taurus moved to Bainbridge, GA in 2019, 'the records from Miami were put on microfiche and we do not have access to those records now' (maker email published 2024-01-31 in th_tips). An earlier maker reply (2015-01-12): 'their records only date back to 81'. A third reply (2022-02-14) says serial-less/number-only handguns 'were sold in bulk, usually to worldwide police departments ... they applied their own serial numbers'.

8 printed anomalies, kept visible because they are the evidence the folklore fails:
  • prefix 'BO or B0' on a brand-new 627 Competition Pro (bought 2025); a reply read it as 'BO = March 2008' under Reading 1 - arithmetically fine, impossible for a model introduced around 2024; no third letter, 'B0' or 'BO' may not even be the same characters (th_tips 2025-09-12)
  • 'My PT92AFSD Serial Number TLA3385S/AFD ... purchased new in 1984 ... mfg in miami starts with I or 1? then e84' - the owner cannot read his own prefix (th_75526)
  • TYE85xxx bought 'in early 2007': fits no printed ladder or month table; owner: '2006 and earlier is a whole new ballgame' (th_22412 2009-07-28)
  • TJY53XXX: 'so the last month chart doesnt work' (th_75526 page 3, 2018-10-13)
  • the forum expert answers 'K179xxx' with 'KI means: K = 1991 I = September' - silently converting the digit 1 into the letter I rather than refusing (th_75526 page 3)
  • '1GA' read by a user as 'First model made in Georgia' (claim), while 1FO serials carry August 2017 grip-stamp dates - three years BEFORE the Georgia move, and 1GC got 'made in Brazil April 2022' from the maker's phone line: the prefix does NOT encode geography as the folk claim says (th_gx4)
  • 'Revolvers has only two letters before the serial number' (th_627 2022-01-06) immediately refuted by the same thread's ACD three-letter revolver and the 2019 'ALL guns' rule (th_627)
  • landshark069's G3C: 'just received Mar 7, 2025 ... frame stamp showing 5 / 22 ... box label ... 25.5.2022' - post timestamps in the imported legacy threads are unreliable (some posts carry pre-import vBulletin dates); forum dates are captured as-displayed and flagged (th_lookup - reported by the thread, not re-located in the 2026-10-21 re-check, so it is kept as reported and flagged unverified)
Month letters that disagree with the guns themselves (why a month answer ships as “the month letter says X”):
  • ABJ ('October, 2020 by this system. Problem is I bought it in September, 2020') — Decodable prefix carrying a printed one-month conflict: the ladder answer is October 2020, the purchase on record is September 2020, and both are printed here - neither is hidden.
  • ACL (frame stamped Oct 2021, box label 21.10.2021 (Brazilian day-first), serial letter says December) — Decodable year (2021), contested month: the serial's third letter says December while the frame and the box label say October 2021. The year answers, the month is never stated as fact, and both readings appear.
  • AHG ('two Taurus revolvers made in July 20 26' from a post dated 2025-08-16) — Either the year letter runs ahead of the calendar year or the owner's date is wrong; the month is not asserted and the conflict is printed here.
What is NOT established (6 items) — the tool refuses these rather than filling them in:
  • pre-1981 numeric serials (any date at all) — No published rule; the digit-count folklore is self-labelled 'not 100% accurate' and revolver-specific; the maker's own records are microfiched or start at 1981; some number-only guns were police bulk sales renumbered by their departments (maker reply 2022-02-14).
  • which factory (Forjas Taurus, Porto Alegre vs Taurus USA, Miami/Bainbridge) issued which prefix — Owners report 'Made in Brazil' on Bainbridge-era guns (phone answer for 1GC) and Georgia-era prefixes on Brazil-marked guns; a 2022 post calls a factory-specific scheme 'I'm guessing'. No printed registry by factory exists. The one-digit 1-series does not track geography either (1FO predates the move).
  • pistol 1981 boundary — The 1981/Nov-1988 introduction dates are stated for revolver frames only; the earliest printed pistol letter scheme in the sources is a 1984-purchased PT92 whose prefix (TLA...) fits none of the ladders comfortably.
  • what the 2019 pair does at Z (2044 rollover) or whether the FIRST letter cycles — All printed examples keep the first letter fixed at A; no source says what happens after AZ.
  • Brazilian domestic-market serials — Only touched second-hand (a Brazilian author's book not available in English; 'European guns follow a different pattern?' asked and unanswered). Out of scope for the US registry.
  • meaning of 'Seventh Edition of official Taurus manual' — The rule is relayed by one forum poster (Erick Tamberg) in two threads - not by two independent posters - and the manual itself could not be reached today on either maker site. The claim is recorded as relayed, never verified against the book.
The guard the tool never crosses. Taurus's own serial lookup is retired: this tool does not proxy it, does not link it as if it still worked, and does not call it available (HTTP 404 on both its URLs, measured 2026-10-21). The serial search under the maker's repair-status page answers repair records only - calling that a dating service would be false. And where the two published charts disagree about your serial, this tool prints both readings instead of averaging them into a compromise year.
Serial numbers are not a factory letter. A year here is a prefix reading from collector ladders the maker itself never published. The month letter can disagree with the gun’s own frame/box stamps (ACL, ABJ, AHG are printed on this page with their conflicts), and the two older-era readings are not reconcilable at all. None of this is a factory letter: the per-serial database that could have been one was retired with the maker’s lookup (dated 404s above), and the records behind pre-1981 guns are microfiche the maker cannot consult.

If your model is not listed, Taurus 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?

The honest reason list for a Taurus is long, and every item on it is sourced, not hedged. The serial is <strong>numeric-only</strong> (pre-1981 boundary: the Miami records were microfiched in the 2019 move, “records only date back to ’81”, and some number-only guns are police-renumbered bulk sales) — <strong>not_established</strong>. The prefix is <strong>two or three letters B-Z</strong> — the two published readings disagree about what the first letter means and the refusal names both, verbatim. It contains <strong>I or 1 (O or 0) in a position that changes the reading</strong> — refused, because even the forum expert silently read K179xxx as “KI” and the 2025 “BO/B0” revolver fits no scheme. It is a <strong>1FO/1GA/1GC</strong> 1-series — printed in the wild, ruleless, phone-answered per serial. It is <strong>AI or later</strong> — beyond the printed ladder, the rollover unprinted. None of these is rounded to a neighbouring year, and no answer here is averaged against a rival reading to “fix” it.

How accurate is this?

The year pair rests on one collector source (Erick Tamberg) reporting the update of Dr. Domingos Tocchetto's manual, plus specimens with independent purchase dates (ABD, ACL, ABJ) — re-opened 2026-10-21 — and zero maker-published rows: an honest collector grade, not a factory record. The AA-AD pairs are printed, AH is observed once, in a report that contradicts itself (its own date predates the year it claims), AE-AG are projections and answer labelled as projections. Month answers ship as what the letter says, because real guns’ stamps disagree with the month letter (ACL, ABJ, AHG — printed on this page). For a definitive date the maker once held the only per-serial answer — and retired it; what survives answers repair status, not birthdays. Of the 265 Taurus rows in this site’s database, 143 carry a year field and 16 carry impossible years beyond 2026 — enumerated, never used as evidence; see the coverage note on this page.

Where does the data come from?

one collector source (Erick Tamberg) reporting the update of Dr. Domingos Tocchetto’s manual, opened 2026-10-21 — Taurus itself publishes none — Zero manufacturer sources today: Taurus retired its per-serial lookup (both URLs dated 404 on this page) and publishes no serial chart — measured 2026-10-21. What is printed here comes from one collector source (Erick Tamberg) reporting the update of Dr. Domingos Tocchetto’s manual — Tocchetto authored the official Brazilian Taurus/Rossi manuals (“ACD means April, 2021”) but reaches this record through that reporter’s posts — quoted verbatim with poster names and dates, plus specimens with independent purchase dates (ABD, ACL, ABJ). AA-AD pairs are printed, AH is observed in the wild, AE-AG are labelled projections. Where the published readings disagree (1981-2006 and 2007-2019) this page prints BOTH, never an average.

Records & archives

A chart says what the factory serials mean. Our records note for Taurus says who actually holds the factory records today, whether a letter of authenticity can be ordered, and what it costs and takes — observed, not assumed.

The Taurus records & archives note →

Other makes

Serial-number charts are published for: