LEGO® minifigure variant differences

≈22 min read · Updated August 2026

Somewhere in your collection right now, there is probably a figure whose almost-twin is worth ten times as much. Or a tenth as much. The difference between them is small enough to cover with a fingertip: an arm color, a back print, the way a stud is molded on top of a head. I've written before about what your minifigures are worth and how to identify them in the first place... this page is the missing middle step. You know WHO the figure is. The catalog is showing you two or three nearly identical entries. Which one is in your hand, and does the difference actually matter?

That's the whole page. It's long, because it's meant to be the reference you come back to, not a read-once article. Use the jump links, keep it open in a tab while you price, and if you only remember one sentence, make it this one: print, mold, and color differences are how you NAME a figure... scarcity of supply is what PRICES it. We'll earn that sentence with receipts below.

The vocabulary, in one minute

Collectors use these words loosely, but they mean different things, and the differences decide what your figure is worth:

• A version is a deliberate redesign of a character. LEGO® made a new Boba Fett with better printing; both are real, both are catalogued, they are different figures.
• A variant is a smaller split inside what looks like the same release: one part's color changed, a print was revised mid-run, a mold was updated. The catalogs sometimes split these into separate entries... and sometimes don't. (The hobby's version/variant vocabulary traces to LEGO®'s own Star Wars character encyclopedia, and the community has run with it ever since.)
• A misprint is a production error in the printing itself: shifted, smeared, incomplete. The classic test: if the print is uniform and clean, it's a variant; if it's misaligned or partial, it's a misprint.
• A factory assembly error is the one everybody forgets, and it's where that classic test gives you the WRONG answer. A figure that left the factory with one arm in the wrong color has a perfectly uniform, perfectly clean print... and it is still an error, not a rare variant. If the “variant” you found is a single part that doesn't match any catalog entry for that figure, think assembly error (or a previous owner's parts swap) before you think treasure.
• And wear mimics all of the above. A worn belt print can impersonate the plain-belt variant. Sun fade can impersonate a color variant. Rule out condition before you declare the rare twin. Seriously.

Three ten-second checks before anything else

Before you compare prints and studs, three quick screens. Each one has ended a pricing session early for somebody.

1. Is it glued? Keychain figures carry a metal ring anchored down through the head stud, and their joints are usually bonded. Magnet figures are sneakier: the ones from around 2009 stood on magnetized bases and come apart normally, but from around 2011 LEGO® started gluing the magnet figures' torsos and legs (collectors documented the switch at the time). A glued figure is catalogued separately, prices differently, and is not the poseable figure your buyer thinks they're getting. Check first. (And no, I won't be teaching you how to un-glue one. That experiment always ends in broken plastic.)

2. Is the print even official? Aftermarket printing houses produce beautiful custom designs on genuine LEGO® blanks. Every mold-era test on this page will pass, because the plastic IS genuine... only the print provenance differs. If the design doesn't exist in any catalog, you're probably holding a custom, and it prices as one.

3. Does the assembly match a real catalog entry at all? Used-market figures get reassembled constantly: heads swapped, arms replaced, a lot's parts shuffled by a kid thirty years ago. The hobby calls the result a frankenfig. A 1984 Classic Space figure wearing the 2019 reissue of its helmet (the reissue adds a small dimple on top... more on that below) is not the 1984 catalog entry anymore. It's a mix of parts, and honest pricing treats it that way. Check the figure against the entry's actual part list, not just its vibe.

How two figures with the same name can differ

Every near-twin pair in the hobby comes down to one or more of seven axes. This is the reference core of the page.

1. Print revisions

The same design, printed differently: a logo simplified, a shirt-tail line dropped mid-run, eyebrows changed from black to reddish brown. Sometimes the catalogs split these into separate entries and sometimes they don't (the rules are below). The gaps are usually modest... in the matched pairs I checked, mid-run print revisions ran around 1.1x to 1.4x at the median. Usually. The 1989-logo Batman runs about 8x its simple-logo sibling (as of August 2026), because the sets that carried it are retired and sought-after. Which is a preview of the pricing lesson: the print identifies the figure, but the print alone is rarely what moves the money.

2. Color changes

One part's color differs between twins: arms, hands, a belt, a head. This is the axis the catalogs respect most (a part color change essentially always earns a separate entry), and it produces some spectacular gaps. It also includes the plastic itself changing under everyone's feet: in November 2003, LEGO® adjusted its core grays (and brown) for the 2004 sets, which is why the hobby speaks of “old gray” and “bluish gray”. That changeover is a genuinely useful dating line... and here's the part almost nobody says out loud: in the matched old-gray versus bluish-gray pairs I checked, the gray swap by itself carried no price premium (three pairs, median 0.98x). The color difference names the era. It doesn't, on its own, name a price.

One more color honesty note: a single figure's parts can come out of different factories, and collectors have documented subtle color mismatches inside one brand-new figure (and even factory-specific mold marks on the sprue points). It's real, it's minor, and it is not a catalogued variant... don't pay a premium for it, and don't panic over it either.

3. Mold generations

The molds themselves changed over the decades: head studs went from solid to blocked to hollow to vented, torso interiors gained ribs, legs gained a seam and later a © stamp under the toes. Two things matter here. First, these are fantastic DATING tells (the sixty-second table below). Second, and this surprises people: BrickLink mostly does not split minifigure entries by mold generation. Your 1980s-stud head and a 2000s-stud head of the same design usually live under one catalog entry with one price. In BrickLink's own words, figure parts with mold variations are “handled in ways that are confusing and sub-optimal”, and mold differences earn a split “only for significant functional or appearance differences”. The plastic knows the era. The price guide often doesn't. The exception that proves the rule: the NBA-era torso of 2003 to 2004 was externally different enough that BrickLink does catalog it separately... which makes it a hard two-year date bracket on the rare day you meet one.

4. Part swaps

The same character, shipped with a different piece: a cape in a different fabric or color, a pauldron added, plain legs versus printed legs, short legs versus standard. The catalogs handle these well (a different included part = a different entry), and some of the hobby's famous twins live here. The clone trooper with pauldron and kama runs about 4x its plain sibling. The dual-molded-legs Han Solo runs roughly 28x its printed-legs counterpart (about $8.50 versus about $241 as of August 2026)... though as you'll see below, that gap has more to do with where each figure was sold than with the legs themselves.

5. Construction

Glued, keychain, magnet, display. Covered in the ten-second checks above; listed here because it is a true axis: the catalogs maintain separate entries for these, and the price difference is severe.

6. Reissue runs

The same figure, produced again years later: sometimes identical, sometimes with a tiny tell (a helmet dimple, ribs inside the torso, a slightly different hat color after a palette change). Reissues get their own section below, because the price behavior is genuinely counterintuitive.

7. Distribution (the invisible axis)

And then there's the axis you cannot see on the figure at all: where it was sold. A convention handout, a store-exclusive pack, a figure that only ever shipped in one poorly-selling set. Physically it may differ from its twin by one belt color. Commercially it's a different universe. Hold that thought for the pricing section.

Reading the catalog's variant language

BrickLink is the hobby's pricing catalog (Brickset displays the same minifigure IDs, by its own description “courtesy of BrickLink”... so no, checking Brickset doesn't independently confirm a split). Its notation is learnable in five minutes and it will save you from the classic misreads. I went deep on part numbers in the minifigure parts guide; here's the variant-specific layer.

Minifigure IDs look like sw0036a: a theme-ish prefix invented by BrickLink, a sequence number that encodes nothing (it's submission order), and sometimes a letter. The traps:

• The letter never tells you WHAT changed. It means “differs somehow from the entry without the letter”. sw0036a and sw0036b differ by head color. sw0179 and sw0179a differ by an eye print. sw0175, sw0175a and sw0175b mix a hair MOLD change and an eye print change in one letter sequence. You always have to read the names.
• No letter doesn't mean “the original” or “the common one”. It usually means “catalogued first”. And some families are all letters with no base entry at all.
• No letter also doesn't mean no variants exist. BrickLink sometimes gives the same class of difference a whole new base number instead. If you want to know whether your figure has documented twins, read the entry's name and its related items, not just its ID.
• “new” on the end means “(Reissue)”, and new2 means a second reissue. You won't find that suffix in BrickLink's published numbering scheme... it's used anyway. cas050 is a classic Ninja; cas050new is its reissue, separately priced.
• “s” on the end means stickers applied. A stickered figure is its own entry with its own price-guide line.
• “Undetermined Type” in a name is a warning label. It means the listings under that entry are NOT variant-specific: you're buying a lottery ticket on which twin shows up.

Part IDs look like 973pb0500c01: base part (973 = torso), print code (pb0500), assembly code (c01: this torso with specific arm and hand colors, which are spelled out in the item NAME, not in the color field). Two traps worth the price of admission: on LEGS the c-number is a COLOR ID, not a sequence (970c11 is “hips and black legs”, because 11 is BrickLink's color number for black), and print codes are per-part counters, so pb0001 on one part has nothing to do with pb0001 on another. And if you're starting from the part number instead of the figure (you're holding a loose printed torso and you know it's 973pb-something): the part's own catalog entry lists every figure it appears in. That list is your shortlist of candidate twins, and the identification guide walks the whole part-to-figure road in detail.

What earns a split, officially. BrickLink's figure guidelines lay out the hierarchy (in a document its catalog team labels a discussion draft, which tells you how settled all this is): part COLOR changes essentially always earn a new entry; print changes earn one when they look intentional on LEGO®'s part; mold differences only for significant functional or appearance changes; stickered figures always. And a matching list of things the catalog deliberately ignores: molded part numbers, the LEGO® wordmark styles on studs, printed-versus-molded renditions of the same design. Until late 2018 there weren't even formal variant records being collected. The catalog is a working tool built by volunteers over decades, not a laboratory taxonomy... treat it with respect and skepticism in equal measure.

And one recent earthquake: in February 2024, BrickLink merged away thousands of part-level mold-variant entries (more than a thousand of them blocked- and vented-stud entries, minifigure heads among them). The blocked-stud and vented-stud head entries you may have bookmarked years ago are gone; those heads now live under the plain entries, with the variant detail demoted to seller notes. Two consequences. If an old guide, tool, or your own inventory snapshot uses pre-2024 head numbers, it now disagrees with BrickLink itself. And a handful of merged entries still resolve because listings reference them... so “the page loads” is not proof the number is current.

Checking a figure in hand: cheapest checks first

Here's the order that wastes the least of your time. You'll resolve most figures at step one or two.

Step 1: eyes only. Front print AND back print (back prints are where revisions hide). Arm, hand, and leg colors against both candidate entries. Leg style: plain, printed, short, medium, dual-molded. Cape fabric if there is one: the traditional starched capes gave way to soft “spongy” stretchable fabric in 2015, and BrickLink names the fabric right in the entries. Flip it over: a © LEGO stamp under the toes says the leg mold is roughly 2013 or later.

Step 2: side-by-side with the catalog photos. This is genuinely how I do it myself when I'm unsure: identify the figure, then compare side by side until the difference declares itself. Pull up BOTH candidate entries and compare against your figure part by part. Read the full entry names out loud if you have to... the difference is always named, even when the ID letter hides it. This is also where you catch the frankenfig: a part that matches NEITHER entry.

Step 3: the interior tells. A head's part number is molded inside, readable through the neck opening once the head is off the torso. Torso interiors have gone through three generations over the decades, and BrickLink formally uses interior ribs as a REISSUE marker on a specific list of classic torsos (the seller note reads “Inside with Ribs (Reissue)”). The full interior story, with photos of what you're looking for, is in the parts guide. What about mold numbers on the outside of parts? The design numbers are documented in LEGO®'s own materials, and I'll be honest with you: I've been told they exist, but I have never been able to see one myself. The plastic names the MOLD, not the figure... it will never tell you which print revision or which catalog entry you're holding.

Step 4: rule out condition. Before you conclude you own the plain-belt variant, make sure you don't own the printed-belt variant with a worn belt. Wear impersonates variants. Fade impersonates color changes. This check is free and it protects you in both directions.

Step 5: when you still can't tell. Sometimes the honest answer is “this entry's listings aren't variant-specific” (that Undetermined Type label), or “this difference isn't catalogued at all”. At that point: photograph the figure well and ask. The BrickLink catalog-identification forum and the minifigure communities resolve mysteries daily, and genuinely uncatalogued finds do surface. A difference nobody can name is not a payday yet... it's a research project.

How to date a LEGO® minifigure in sixty seconds

This is the lookup card. Every date below is an era boundary, not a light switch: molds overlapped for years, old stock kept shipping, and the transitions were gradual. (The deeper story on most of these, including what the insides look like, lives in the parts guide.)

One warning on this whole table: a single mold tell BOUNDS a date, it doesn't pin one. LEGO® has shipped two mold variants of a part in the same set. Stack tells before you conclude.

Where the money actually lives

Now the part you came for. I checked dozens of documented twin pairs against live sold-data-derived values for this page (all figures below as of August 2026, BrickLink-derived values via Brickset). Here's the honest shape of it:

Decoration differences are usually modest. Matched print-revision pairs: around 1.1x to 1.4x at the median. The black-versus-brown eyebrow revision across seventeen matched pairs: a 1.35x median that flips DIRECTION in five of them. Printed versus plain legs: about 1.5x at the median. The old-gray to bluish-gray swap: no premium at all in the pairs I checked. A mold revision BrickLink split into its own entry (a cape cut change on a Batman): 1.12x, which is to say, the catalog cared and the market didn't (and that entry's two sets were promotional runs, so read even the 1.12x gently).

The order-of-magnitude gaps all trace to distribution. Same Batman suit, same cowl, different belt color: about $6 versus about $123... and the $123 one shipped only in one tiny 63-piece set... a set that now trades for roughly the price of the figure inside it. Same Kai, one armor piece different: about $9 versus about $176... the expensive one came in a retailer-exclusive promo pack. The famous Cloud City Boba Fett (first printed arms, 2003): Brickset's own editors note it's the most expensive figure from a regular set because the $100 set sold poorly. Even within one figure: a 1978 Classic Space white spaceman, its 2009 limited reissue, and its 2010 wider reissue... the 2009 reissue holds parity with the 1978 original, the 2010 one sits at a fifth of both. Nearly five-x between two REISSUES of the same figure, separated only by a logo position and how many sets they shipped in.

So here is the sentence again, now with its receipts: print, mold, and color differences are how you NAME a figure; scarcity of supply (an exclusive channel, or simply how few sets it shipped in) is what PRICES it. Decoration axes cluster under about 1.5x at the median. The 5x-to-20x gaps in this guide all trace to distribution.

Some verified pairs to make it concrete. A note before you read them: some of these differences hide under a fingertip, and others you can see across the room... but in a listing title with one photo, every pair below collapses into the same name (values as of August 2026; check the live price guide before acting on any of these):

Notice the Batman row. The GOLD belt is the cheap one. The plain yellow belt is the $123 figure. The fancy-looking twin is not automatically the treasure... the trap runs in both directions, and the only way to know is to resolve the entry and read the sold data.

Three honest cautions before you price off ANY premium, including the ones above:

1. Check how many actually sold. A value derived from two sales is a rumor with a decimal point. Look at the times-sold count on the price guide, mind the gap between asking prices and sold prices, and know whether the number you're reading is an average or a median (one outlier sale drags an average a long way).
2. The price guide inherits everyone's mistakes. Sellers mislist twins under the wrong letter constantly... which pollutes the very sold-data that defines the premium. (That circularity is half the reason this page exists.)
3. The bigger the premium, the more the one different part matters... and the more it attracts fakes. Brickset's editors put it perfectly about a convention-exclusive Yoda that trades in the thousands, whose torso is the only unique part: it's “easily counterfeited”, so serious buyers want it sealed with paperwork. Mr. Gold (col161, the Series 10 chase figure, 5,000 made) shows values around $8,300 new and about $1,525 used as of August 2026... and the same editors note “there are a lot of fakes around”. A one-part difference is exactly the difference a faker can manufacture. High-premium twins deserve the authenticity checks before the celebration.

(The full pricing framework these cautions live inside, venues and all, is the worth guide.)

And one for the collectible-minifigure folks: I went looking for same-character CMF twin pairs with real variant-driven price gaps, and they essentially don't exist. The CMF value story is print-run rarity and series, not variants. If you're pricing CMFs, your questions are “which series” and “is it complete”, and the worth guide has you covered.

Reissues: same figure, second serving

LEGO® re-releases classic figures more than people think, and the catalogs mark it in about four different ways (of course they do):

• A separate entry with “new” or “new2” in the ID and “(Reissue)” in the name.
• A seller note: certain classic torso designs reissued with modern interiors carry BrickLink's “Inside with Ribs (Reissue)” annotation.
• A physical tell molded in: the 2019 reissue of the beloved Classic Space helmet added a small dimple on top; a 2014 commemorative version even molded in a simulated crack (a wink at every vintage helmet that cracked at the chin strap).
• A letter variant with the year in the name, when a rerun's print changed subtly.

What does a reissue do to the original's value? Less than you'd fear, and less predictably than you'd hope. In the standings I checked (as of August 2026): several vintage figures sit at rough PARITY with their limited-pack reissues; the Blacktron spaceman's reissues actually price above the 1987 original; and the widest gap I found between reissues of one figure (that 5x Classic Space case above) came down to distribution width, not the reissue-ness itself.

The scary story is the other direction, and I told it in the worth guide: Captain Rex, sw1315, shipped exclusively in a $650 set in October 2023 and traded north of $100... then in June 2024 the same catalog entry shipped in a $12.99 set, and the price collapsed under $10. Not a variant, not a reissue entry: the same figure, suddenly everywhere. That's the supply thesis with teeth.

Three dozen Vaders walk into a search box

Last trap, and it's the marketplace itself. BrickLink catalogs three dozen distinct Darth Vaders. A marketplace listing titled “LEGO® Darth Vader minifigure” could be any of them, spanning more than an order of magnitude in value. So:

• When you read listings: never buy a premium twin from a title. Make the listing name the entry ID, or make the photos show the distinguishing part (the back print, the arm color, the cape fabric). “Undetermined Type” entries are lotteries; price them like lotteries.
• When you write listings: pin the ID and photograph the tell. List the wrong letter and one of two things happens: a knowledgeable buyer disputes it, or a wanted-list buyer never sees it at all (a buyer watching for the “b” variant is never shown your mislabeled “a”).
• When you sort a lot: LEGO® has shipped BOTH twins in the same set number (BrickLink documents hundreds of in-set variant cases), so “I already checked one of these” is not checking. In a bulk lot, near-twins hide in plain sight next to each other. Check every figure that has a documented expensive sibling.

The FAQ corner

What's the difference between a minifigure variant and a version? A version is a deliberate redesign of a character; a variant is a smaller split inside what looks like the same release (a part color, a print revision, a mold or included-part change). The catalogs give both their own entries when the difference is judged intentional and significant... and the entry NAMES spell out what differs, even when the ID doesn't.

How do I tell which variant of a minifigure I have? Read both candidate entries' full names, then compare your figure part by part against both sets of catalog photos: front and back prints, arm and hand colors, leg style, cape fabric, included parts. Most twins resolve with eyes only. If the difference is interior (head numbers, torso ribs), the checks in the parts guide go deeper.

Do arm and hand colors really matter? Yes. A part color change is the difference the catalogs respect most, and some of the hobby's biggest twin gaps are a single limb color. Always check limbs against both entries before pricing.

My figure's print is misaligned. Is it a rare misprint worth money? Usually no. Clean, uniform print = variant; shifted or incomplete print = misprint, and most misprints are curiosities rather than premiums. A uniform print in the WRONG color on one part is more likely a factory assembly error or a parts swap than a variant. Documented misprint sales exist, but they're the exception... verify against sold data before celebrating. (And if the real story is a swapped part, the repair guide covers restoring the right one.)

How do I date a LEGO® minifigure? Stack the era tells: head stud type, torso interior, leg seam and under-toe stamp, skin tone, gray shade, cape fabric, leg style. The sixty-second table on this page brackets most figures to an era in about a minute; no single tell is a precise date.

Are old gray minifigures worth more than bluish gray? By themselves, no. In the matched pairs I checked, the November-2003 gray changeover carried no price premium on its own (median 0.98x across three pairs). It's a dating tell, not a payday. The valuable “old gray” figures are valuable for their scarcity, not their shade.

What does “(Reissue)” mean on a BrickLink minifigure entry? LEGO® produced the figure again years later, and the catalog split the rerun into its own entry (IDs ending in “new” or “new2”). Reissues price independently: sometimes below the original, sometimes at parity, occasionally above it.

Why did my saved head part number stop working? In February 2024 BrickLink merged thousands of mold-variant part entries, including more than a thousand blocked- and vented-stud entries, minifigure heads among them. Those head numbers were folded into the plain entries, with mold detail moved to seller notes. Anything that snapshotted head numbers before then now disagrees with BrickLink itself.

What is a frankenfig? A figure assembled from parts of different figures... deliberately or by a kid in 1993. It matches no catalog entry exactly, and it prices as parts, not as the entry it most resembles. Part-by-part checking against the catalog reference is how you catch one.

Can I tell variants apart without taking the figure apart? Almost always, yes: prints, colors, leg style, cape fabric, included parts, and the under-toe stamp are all exterior. The head number does ask you to pop the head off and read through the neck opening (gently... that part is routine handling). The rare exceptions (torso interior generations that BrickLink never split) mostly matter for dating vintage reissues... and for those, the ribs note and the era table above get you the rest of the way.

Thank you for reading this far... genuinely. This page exists because the wrong-twin trap is the easiest expensive mistake in this hobby, and I wanted the reference I kept wishing existed while I sorted. If Figventory can eat some of that work for you, wonderful (the app is here, free tier and all)... and if all you take away is the habit of reading both entry names before you price, that's a win I'll happily take. Zero pressure. Happy sorting.

— Matt