The Commonwealth of Kentucky first registered motor vehicles in June 1910. Porcelain enamel plates were issued bearing no date; just the letter "B." Three more series of plates followed, each of them bearing a cryptic letter code: "L," "M," and "G." Finally in 1914, the state transitioned to annual dated plates, putting closure to an era that would intrigue and confuse collectors for the next century and then some.

The order in which the letters were used isn't in doubt. Nor is it a mystery why the state chose to use letters: Kentucky plates were valid for one year from the day of issue, and could expire on any day of the year. New designs were introduced in June, and coexisted with the old until all registrations had come up for renewal. Any form of visible dating would have been counterproductive, since some "B" plates were valid in 1910-1911 while others were valid in 1911-1912. So arbitrary identifiers for each registration period were used instead.

What isn't known...and isn't likely to ever be known with certainty...is why the state chose to use the letters that it did. As far as I know, the Kentucky laws and surviving records of the period either make no reference to the "B," "L," "M," and "G" letters, or refer to them matter of factly without elaborating on them in any way.

This plunges us into the realm of speculation. And although it's tempting to attribute "B," "L," "M," and "G" to elaborate schemes or cryptographic codes, Occam's Razor holds that the simplest explanations are the most likely:

• "A" would have been a logical starting point for the sequence. Kentucky bumped it by one to "B," perhaps to avoid vowels or to avoid "KYA" being interpreted as an erroneous abbreviation.
• "L" and "M," used primarily in 1912 and 1913, are respectively the 12th and 13th letters of the alphabet.
• "G" appears to be an arbitrary letter chosen completely at random. Perhaps the state became concerned about the sequence of letters being linked to the calendar year, especially since delays resulted in the "G" plates debuting over two weeks late.

What a 1905 WV plate MIGHT look like.

West Virginia began licensing the operators of motor vehicles in May 1905. (Until 1917, all West Virginia licence plates identified motorists who owned vehicles, but not the vehicles themselves!)

No genuine West Virginia licence plates dated 1905 are known to exist. The earliest surviving plates are dated 1906. Over the years, several explanations have been proposed to explain this:

• Explanation 1: The 1906 issue was introduced in 1905, and the "1906" date reflected the expiration year.

  This was cited by Jim Fox in the 1990s. Except...Eric Tanner says that 1906-dated plate 209 survives with its accompanying licence certificate, issued in 1906 to expire in 1907!

• Explanation 2: A 1905 issue existed: A tiny 3"x5½" plate marked "WVA 05." Only one survives.

  The plate in question was "discovered" by Shib Pixley and Howard Platte in the 1950s, and looks like it was embossed with a screwdriver. ALPCA's official word is that it's a fake: "This plate was made by a very early ALPCA member, who acknowledged its origin. Over the years, myth replaced fact, as some people sought to legitimize this forgery."

• Explanation 3: A 1905 issue existed, but it's unlikely that any survive.

  One legitimate 1905-issued West Virginia licence certificate survives, all but proving that some plate must have been issued that year. Could an authentic plate turn up? Maybe...but several strikes are against it. Only 172 licences were issued in 1905...a number on par with the rarest U.S. plates, 1913 Mississippi and 1921 Alaska. The 1905 West Virginia plate likely shared the 1906 issue's unusually small size and fragile construction, making it easily lost or damaged. It could have also been issued singly, reducing numbers even further. The 1906 issue survives chiefly through a cache of unissued plates found in the State Capitol Archives...but no such cache exists for 1905.

Mexico began issuing nationally-consistent licence plates in 1933. All-numeric configurations were used until 1939, when increased quantities resulted in "A" and "B" suffixes as an overflow series. To prevent future overruns, Mexico's numbering system was thoroughly overhauled in 1940, with all passenger vehicles adopting an alpha prefix format.

The consensus among collectors is that these letters are mnemonics tied to the state of issue...much like the systems in use from 1947 to 1953 and 1960 to 1965, when state names appeared on the plates. Unfortunately the plates of 1940 to 1946 lack such embellishment, making verification difficult. Nor is any readily-available reference material known to exist.

Extrapolating the 1947-1953 letter blocks to the preceding years doesn't work because of numerous discrepancies. The older system makes use of I, R, and W...letters that aren't correlated with Mexican state names at all! Similarly, the newer system uses letters like G, H, and O that don't appear consistently prior to 1947:

As you can see, the 1940-1946 letter blocks are as clear as mud. Early Mexican plates aren't exactly common in the collector market, either, making it difficult to accrue observations that might make the system come into clearer focus.

It's possible that the answers are still out there, lurking in dusty ledgers or microfilms somewhere in the Archivo General de la Nación. Finding them would require international travel, time, and fluency in the Spanish language.

In 1929, California overhauled its licence plates with a format, die style, and colour scheme freely copied from New York. Plates followed a 1A-12-34 format, allowing for more than 2 million combinations without the need for overflow formats.

With letters, however, came confusion. Many collectors swear that California plates of this era were county-coded. Blocks of prefixes were undoubtedly assigned and distributed on a geographical basis...but the prefix assignments weren't directly correlated to counties, and they changed and shifted from year to year.

What's more, official records of the allocations don't exist. What's known about them is piecemeal and incomplete, drawn mostly from old newspapers (courtesy of Eric Tanner):

In the early years, many of the allocations had gaps in sequence between them...making it difficult to figure out how many plates were made by serial ranges alone. By the late 1930s, however, the 1A-23-45 sequence had been filled to the brim and California was delving into overflow formats. Sequencing got even more squirrelly in the 1940s thanks to extended baseplate use. And where were the geographical allocations caught in the middle of this? No one knows.

When a distribution office depleted its original allocation of plates, did they receive plates from a single "catch-all" series, or did the state attempt to impart a geographical correlation to these allocations as well? No one knows.

California's 1951-55 baseplate was the last gasp for the remnants of the 1929 serial format, and it brought the state's serial numbering to its longest, most confusing heights.

Accurately plotting the serial progression of these plates is almost impossible. There are several exacerbating factors at play:

• Plates were distributed and issued at the outset in geographically-correlated allocations. As with other California issues back to 1935, most of these allocations are unknown. The few that are known are all over the map: Fresno was allotted three different starting points of 8B 2001, 5K 1001, and 9B 10000!
• With the advent of this baseplate, serial formats were changed to prioritize a seven-digit configuration and eliminate prefixes of two numerals and one letter. This meant that the precedents of prior years no longer applied.
• Plates were validated with easily-removable metal tabs, not stickers. This makes it impossible to gauge the issue year of a plate from the strata of its sticker stack.

The following is known about the plate numbers that were issued:

• 5- and 6-digit combinations (i.e., 1A 001 or 1A 1234) were used to exhaustion, or nearly so. Skipping I, O, and Q, every prefix from 1A through 9Z has been seen in this format, with one exception (3N).
• 7-digit combinations were issued early on in the 1A-5A, 1B-9B, and 1N-9N series.
• Once the original geographically-correlated allocations began to be depleted (possibly before the end of 1951), subsequent plates appear to have been doled out of at least two separate series that progressed in parallel, filling in the gaps of the 7-digit format with the prefix letter progressing before the prefix digit.
• Despite this, a certain proportion of 5- and 6-digit plates continued to be issued in 1952 and beyond. Were these from small DMV offices still slowly working through their original allocations of plates? Maybe. Maybe not.
• For some reason, the state issued a limited number of six-digit plates containing the previously-skipped letter "Q" in 1955.

The following prefixes have been confirmed on 7-digit plates:

The key to the mystery might lie in a complete list of original plate number allocations from 1951, or any prison production records that might exist. At least, this would give some way of positively knowing which plates were ready for issue in the beginning, and which were later additions. But if these pieces of information exist, one would think that they would have surfaced by now.

Credit is due to Scott Broady for first broaching the 1951 California mystery in the December 2004 issue of the ALPCA magazine, and Eric Tanner for furthering the search and gathering known evidence on his site.

Pennsylvania's plate history is a litany of gold and blue, keystones, maps, and few surprises. There's one glaring exception: The 1970 "interim" plate, which defies all logic. What could be made of this?

In 1970, Pennsylvania was on the eve of a general reissue. Consensus holds that the "interim" plate was made and distributed on a very limited basis after production of the outgoing 1965 base had stopped, and before the forthcoming 1971 base could legally be displayed. The design was an amalgamation of three different designs, sharing aspects of each:

• The wide dies and validation sticker of the outgoing 1965 passenger baseplate.
• The distorted border die, sticker box, and state name placement of the concurrent 1968 commercial and trailer baseplate.
• The gold-on-blue colour scheme of the forthcoming 1971 passenger baseplate.

But why does the plate look like this? If the rationale for the interim plate was to produce a limited number of old-style plates after the plate shop had retooled, why weren't the outgoing blue-on-gold colours retained...especially since most non-passenger plates retained the old colour scheme for another year?

Also inexplicable is the numbering. If this were a continuation of the 1965 base, the serial range should have been a continuation as well, picking up beyond where it left off around 63U-000 or so. But this isn't borne out by evidence. Three interim plates have been reported: 100-428, 100-537 (pictured on Papl8s.com), and 7R5-501. The first two represent the start of a new series, while the last looks like a remake of a number from 1968.

Could it be that our understanding is all wrong? Maybe the "interim" plate wasn't part of the 1965 base at all...it was a variant of the 1971 base that was distributed early and replaced quickly. This jibes with the colours and numbering. This would also suggest that the numbers on the "proper" 1971 baseplates pick up where the "interim" plates leave off.

Have any "Bicentennial State" plates been verified with a number of 100-537 or lower?

(from various eBay auctions)

From 1955 to 1984, the state of Kansas issued unusual Antique plates serialized with letters instead of numerals. The serials can consist of anywhere from one to seven letters. The sequencing is perplexing, and has never been completely deciphered.

Making matters worse, motorists could also request personalized combinations that weren't part of the sequence at all. It's obvious that word plates like "MOVE," "SMILEY," or "JLOPY" are personalized. But what about "HF" or "MLE?" Many personalized combinations were simply peoples' initials...and these are impossible to tell apart from sequential combinations without additional evidence.

One saving grace of these plates is that they were produced in multiple die variations, correlated to changes affecting Kansas passenger plates over the same era. Here is the predominance of format length by variation, taken from a survey of more than 150 examples:

What to make of this? Generally, the more letters, the later the serial is in sequence...so far, so good. But three- and four-letter plates appeared from the outset, long before the formats should have been required. There surely weren't 18,279 (26+26²+26³+1) antique vehicles in Kansas in 1956!

It's indubitable that not all combinations were used. A recurring theme of these plates are combinations with repeating letters unlikely to be initials. Here are the patterns I've seen:

• AAB (first seen on 1960-68 die variation)
• AAAB (ditto)
• ABBC (first seen on 1968-72 variation)
• AABC (first seen on 1972-75 variation)
• AAAAB (ditto)
• AAABA (ditto)
• AABAA (ditto)
• AAAABA (first seen on 1975-82 variation)
• ABCC (first seen on 1982-84 variation)
• ABAAAA (ditto)
• ABBAAA (ditto)
• ABBBAA (ditto)
• ABBBBCD (ditto)

Most of these configurations allowed for up to 676 combinations apiece (or more), which keeps the number of plates issued to a more reasonable number. Other patterns may also exist.

Other clues to Kansas' sequencing are error plates produced in the 1960s with the "Midway U.S.A." slogan (I've seen both MCB and MWM in this format), and consecutive plates found together: QQQYQ and QQQZQ; KXXXXPC and LXXXXPC.

Despite these leads, we still lack a thorough understanding of Kansas' sequential letter system. The exact order of patterns (and the sequencing within them) remains unknown. Unless production records exist, the only way to solve the mystery is by gathering more evidence through either plates or photos...and to keep an eye out for consecutive plates!

1980 was a watershed year in Ohio registration history. The state engaged in a general reissue and replaced a creaky, duplication-strewn numbering system dating back to 1936 with a brand-new system more suited for computerized records. Passenger plates now had an easy-to-read ABC-123 format. Most non-passenger plates contained a prefix character keyed to the type: Motor Home was E, Farm Truck was F, and Non-Commercial Truck was N, for example.

Naturally, motorcycle plates followed suit. Due to design constraints, however, motorcycles were limited to five characters. This meant that Ohio would go through a bewildering number of alphanumeric configurations until the next reissue rolled around, with anywhere from one to three letters and one to four numerals in seemingly any position! How to put these in order?

In the 2010s, I began taking notes on Ohio motorcycle plate serials. I immediately noticed that all plates regardless of format contained the letters M, Y, or Z in either the prefix or suffix position. I then counted off at least 31 different series:

Ohio plates of this era were pro-rated to different expiration months by initial, making the stickers minimally useful for dating. Nevertheless, with enough samples, some chronological patterns emerged. The M-prefix formats were almost certainly among the first issued, followed closely by the Ys in late 1980 and the Zs in 1981. Suffix formats didn't appear until 1983 at latest. And the A00AM series may have been the last in line, since the only plates seen in that format were unissued leftovers.

It was impossible, however, to deduce any granular conclusions. Did the serial letter float from left to right (YA000, Y0A00, Y00A0, Y000A), or were the series run through in a different order? Did the state deplete all M combinations before Y and all Y before Z, or was there mixing and matching as the state added more alphanumeric configurations to the mix? Did a series like Z0A00 progress right-to-left (so that Z4L12 came before Z5K12), or progress so that numerals advanced before letters regardless of position? For all of these questions, the answer was "No one knows."

Short-format plates were another complication. Was a number like Y5L1 issued as part of the Y0A00 format, or Y00A0? Again: "No one knows."

As of 2026, I'm still slowly parsing away at the 1980-85 Ohio motorcycle format, gathering more evidence through plate pictures and notes, and waiting for a pattern to emerge! While I still don't have the answers, I have happened upon new takeaways and moments of encouragement:

• Following the precedents of other non-passenger types, the M0000 configuration should have been the starting point of the whole affair. Yet, the earliest example I've confirmed had a 2-82 sticker. But there's an explanation for this: This configuration could have been held from issue until 1981 because it conflicted with outgoing 1976-80 plates. Ohio did exactly this for passenger plates with numbers below CAA-010, so the precedent is there.
• Ohio almost certainly engaged in "number control" on some 1980 non-passenger formats: Non-Comm plate N8F608 is a later base variation than N4X534, for example.

• Two state name legend dies were used in this era: One wide, one narrow. Both variations appear scattered throughout the entire series. However, careful observation of these die variations could reveal whether two plates of different configurations were produced as part of the same manufacturing run!