How old is my motorcycle?
If it was built from 1981, this is already answered — free, officially, and on this site. A motorcycle carries a 17-character VIN under the same federal rule as a car, and the VIN decoder here reads it and checks it against the US government's own database. Nothing on this page replaces that. Start there.
This page is about the two things that page cannot tell you: why a motorcycle's year character is more ambiguous than a car's, and what happens when the bike is older than the standard.
The tie-break a car gets, and a motorcycle does not
The tenth character of a VIN is the model year, and the code repeats on a thirty-year
cycle. K is 2019 — and 1989, and 2049. On its own it cannot tell
you which.
For a car, the regulation supplies a way out. 49 CFR 565.15(b) requires that — quoted exactly —
“For passenger cars, and for multipurpose passenger vehicles and trucks with a gross vehicle weight rating of 4536 kg (10,000 lb) or less, the fourth character (position 7) of this section shall be alphabetic.”
And the note under the year-code table then uses that:
“For passenger cars, and for multipurpose passenger vehicles and trucks with a gross vehicle weight rating of 4536 kg (10,000 lb) or less, if position 7 is numeric, the Model Year in position 10 of the VIN refers to a year in the range 1980–2009. If position 7 is alphabetic, the Model Year in Position 10 of the VIN refers to a year in the range 2010–2039.”
Read the list of vehicles in each. Motorcycles are not in it. They are not an oversight either — the same regulation names them explicitly where it says who it applies to, and gives them their own line in the table of what a VIN must reveal: “Motorcycle: Make, type of motorcycle, line, engine type, and net brake horsepower.” No model year, because the model year lives in position 10.
| Vehicle | Position 7 must be alphabetic? | So the 30-year cycle… |
|---|---|---|
| Passenger car | Yes | splits cleanly into 1980–2009 and 2010–2039 |
| Light truck or MPV (≤ 4536 kg) | Yes | splits cleanly |
| Motorcycle | No | does not split. The character means either year. |
This is an argument from an absence, so it is checked as one.
tools-parse-cfr565.py pulls the list of vehicle classes out of both rules and
compares them, rather than searching for the word “motorcycle” and concluding from
silence — a search that finds nothing is a claim about the search. If NHTSA ever add
motorcycles to either list, that tool stops and this section comes down.
This one runs on your own device. It is worked out in your browser, so there is nothing to send and nothing to log. Where a tool on this site does send something out, that page says so and names where it goes.
What actually settles it is the same thing that settles everything else about a modern VIN: the manufacturer's own filing with NHTSA. The VIN page looks that up, shows every year the character allows, and says so when the two disagree.
Before the standard, and Harley's own answer
The 17-character VIN is a requirement, not a description of every motorcycle ever made. Put
a pre-standard number into the government's decoder and it does not guess — it returns
Incomplete VIN and stops. There is nothing behind it to look up.
So what do the makers themselves hold? The most-searched marque on the list answers that directly, on its own website, and the answer is the most candid thing any manufacturer says on this site:
“Can you help me find the history of my motorcycle? Documentation and details around Harley-Davidson production vehicles are very limited or non-existent.”
“Can I do research in the Harley-Davidson Archives? No. The Harley-Davidson Archives is a private, corporate archives and not open to researchers.”
Every other family swept for this site has makers who publish nothing. This is different in kind. It is a manufacturer saying, in its own FAQ, that the records may not exist — and that the archive which might settle it is closed to the people who would look. Watch makers keep the answer and sell it; Baxi keep it and will tell you if you ask. Harley are saying there may be no answer to keep.
That is worth taking at face value rather than reading as a brush-off. A company founded in 1903 that went through the Depression, two wars and the AMF years has no particular reason to have kept a build record for every frame, and saying so plainly is more useful than a lookup form that returns nothing.
What Harley do publish, and a trap in it
Their Service Information Portal carries a genuine VIN breakdown, free and open, which is more than most makers on this site manage. It confirms where to look — “The full 17-digit serial number or Vehicle Identification Number (VIN) is stamped on the right front of the frame near the steering head” — and then breaks the number down.
Their table numbers fields, not characters. It lists nine numbered rows, and row 6 is the model year. A reader who counts six characters into the VIN lands on the wrong one: row 1 is the three-character manufacturer code, row 2 is a two-character type code, and by the time you reach row 6 you are at character 10 — which is where the federal standard puts the year anyway. The table is right. It is just not a character map, and nothing on it says so.
One page of that portal answered us. The others return HTTP 200 with a body reading “Access Denied” — a refusal dressed as a success, which is worth naming because a checker that only reads status codes records it as working. Ours did, once. It now looks at the body.
One real decoder, and the sweep found it
S&S Cycle publish a complete year-code table, free and open, on their own engine serial lookup page. It is the first decoder on this site that the automated sweep turned up by itself — Zippo, Ludwig, Rheem, Trane, Winchester and GE were all found by hand after the same method had missed them. Eleven families in, it worked.
It dates the engine, not the motorcycle. S&S build engines that go into Harley-Davidsons and Harley-based customs, often years after the frame was made and sometimes as a replacement for an older one. The same caution applies as on the amplifier page, where a pot code dates the potentiometer.
The year it gives you is not a calendar year, and S&S say so plainly:
“Year codes coincide with S&S Cycle's fiscal year, which starts October 1 of the previous calendar year. A product with serial number beginning with the letter ‘A’ was manufactured in fiscal year 1978, between 10/1/1977 and 9/30/1978.”
Three months of every S&S year fall in the previous calendar year. Every other scheme on this site is a calendar year, and a decoder that quietly treated this one the same would be wrong about roughly a quarter of all S&S engines — always in the same direction, and never visibly.
Their letters run on a 22-year cycle, because four letters are
marked Not Used: I, O, Q and V.
The federal VIN standard leaves out I, O, Q,
U, Z and 0. Both are solving the same problem —
characters that are too easily misread — and they reached overlapping but
different answers. V is fine in a VIN and unused by S&S;
U and Z are the other way round. There is no standard for what
counts as confusable, only conventions, and this page shows both rather than picking one.
Two letters carry three years rather than two — A and
B — because the third cycle has reached them and no further. How ambiguous
your serial is depends on which letter it starts with, which is exactly the sort of
irregularity a decoder gets wrong by assuming things are tidy.
Their table is not reproduced here, and there is no clever reason why not. This site's rule is to ask a maker before republishing their figures. S&S's own terms page renders in JavaScript and returns eighty characters of navigation to an automated reader, so what it permits is unknown — and an earlier draft of this page carried the whole table anyway, with a decoder built on it. Unknown is not yes. Their page is linked above, it is free, and it is the right place to read it.
It should be an easy letter, and it is worth saying why here rather than only in a file nobody sees. They publish the table in order that it be used. Their own lookup asks for make and model alongside the number, so it answers what a table alone cannot, and this page sends people to it. And what this page adds is a correction in their favour: a reader who takes a bare year off that table is wrong about a quarter of engines, because the year is a fiscal one. If S&S say yes, the decoder comes back.
What was checked
The pool is 649 motorcycle makers, listed in motorcycle-makers.txt, built from Wikipedia's motorcycle manufacturer categories for the world, the US, Japan, the UK, Italy and Germany, plus the defunct ones — with websites taken from Wikidata's official-website property rather than guessed at.
| Outcome | Makers |
|---|---|
Site swept — robots.txt read and honoured | 195 |
| Refused us, or could not be reached | 103 |
| No official website recorded at all | 351 |
And read the swept figure knowing how much was there to read. Of the 195 counted as swept, 83 offered this method fewer than 25 pages to look at, and 65 of those offered nothing from a sitemap at all. A host that answers, allows us, and then has almost nothing to search still counts as swept. It is the number of doors opened, not the number of rooms searched.
That middle column is not the interesting one here. The right-hand one is. 298 of 550 makers — more than half — have no website recorded on Wikidata at all, by far the highest proportion of any family swept for this site. Motorcycle manufacturing has a very long tail of marques that stopped trading decades before the web existed, and the Wikipedia categories collect them faithfully. That is a fact about the data rather than about the companies, and it is the reason a sweep of this family tells you less than a sweep of appliances.
Of the 103 not probed: 59 did not answer at all, 16
returned 403 on robots.txt itself, 17 read fine and then
refused the page, 7 block us by name, 3 went over our
45-second budget for a single host, and 1 disallows everybody.
Read the swept figure as swept, not as cleared. This method has demonstrably missed makers who do publish, on Zippo, Ludwig and Rheem. Harley's own VIN table was found by hand, not by the sweep — it lives on a service subdomain, which is exactly where Zippo's was. S&S Cycle, above, is the other way round: the first one the sweep found on its own.
Eight pages came back worth opening, and five were the word
“series” in another language. The sweep matches serien
because a German maker writes Seriennummer, and Serien alone is the plural
of Serie: so Brembo's Serienbremsen (standard brakes) came back, and so did
Italian and German press releases from Benelli and Brixton. Thunderbike's hit was
Serial 1, which is the name of Harley's electric bicycle brand. That is the right way
round and deliberately so: a false candidate costs one look, and a missed maker is
invisible.
The three real ones were S&S Cycle, above; Honda Marine, whose serial page is about outboard motors rather than motorcycles; and Garelli, whose page is titled Ciclone serial number position and tells you where to look rather than how to read it — and whose content did not render to us without JavaScript, so we are not going to describe what is on it.
Building the list
Makers go in by a positive test: a company is filed on Wikipedia under a category containing “companies”, “manufacturers”, “corporations”, “businesses” or “brands”. Anything with none of those is dropped and reported. 124 entries went that way, and the motorcycle categories collect a remarkable amount that is not a motorcycle maker: the Suzuka Circuit, Yamaha Stadium, VTEC, three video games, a novel series, a lawsuit and several founders.
Two Wikipedia categories that do exist were deliberately not used: Scooter manufacturers and Moped manufacturers. A scooter is a different question from a motorcycle for this purpose and the pool is already 550. That is a decision rather than a gap, which is why it is written here — the tool that builds these lists used to file every name it checked under “does not exist”, which was true by luck on earlier lists and became false on this one.