Credit Card Dimensions: The Numbers Designers Actually Need
You can get wildly creative with finishes, foil, textures, even weird translucent substrates, but if the card doesn’t fit, none of that matters. Wallets don’t negotiate. ATM throats don’t care about your brand guidelines. And printers definitely won’t “make it work.”
So start with the hard geometry.
The core spec (tattoo this on your brain)
The globally dominant credit-card size is ISO/IEC 7810 ID‑1:
– Width: 85.60 mm
– Height: 53.98 mm
– Thickness: 0.76 mm
That’s the “standard credit card” people mean 99% of the time when asking what are credit card dimensions.
If you’re working in inches (and plenty of North American vendors still do), you’ll see it quoted around 3.370 × 2.125 in, and thickness around 0.030 in.
One of those numbers is a rounding compromise, by the way. A lot of US production talk defaults to 3.375 × 2.125 in because it plays nicer with legacy tooling and shop habits. If your supplier says “CR80,” that’s usually what they mean. Close enough for casual conversation; not always close enough for tooling.
Tolerances: where “almost” breaks stuff
Here’s the thing: the card itself is simple; the ecosystem isn’t.
A card gets:
– cut (or punched),
– laminated,
– printed,
– sometimes embedded with a chip module,
– sometimes finished with overlay, varnish, or tactile layers,
– then shoved into machines that assume the world is consistent.
A fraction of a millimeter can move you from “smooth insert” to “why is this jamming in the field?” I’ve seen perfectly good artwork die on the hill of sloppy finishing.
Now, the actual tolerance bands depend on the relevant ISO clauses and your manufacturing method. But your workflow should assume:
– Nominal dimensions are the design target
– Permissible variance is a manufacturing reality
– Critical zones (chip location, mag stripe relationship, edge-to-print alignment) are where you clamp down hardest
One-line truth: tight tolerances cost money, loose tolerances cost reputation.
Corners: small radius, big consequences
Yes, corner radius matters. More than you want it to.
Rounded corners reduce snagging, reduce edge delamination, and generally make cards feel “finished.” Go too sharp and you’ll see wallet wear and corner whitening faster. Go too round and you can annoy certain feeders and stacking systems that “expect” a particular corner geometry.
ISO/IEC 7810 ID‑1 uses a standardized corner radius (the famous rounded credit-card corner). If you’re doing anything meant to behave like a payment card, stick to the standard radius instead of improvising.
And don’t treat it as decorative. It’s mechanical.
Flat edges vs. soft corners (my biased take)
Flat corners look “modern” in renderings. In real life, they’re abrasive little liabilities.
Rounded corners:
– glide better in wallets,
– chip less,
– feel more premium in-hand.
If you’re making a novelty card that never sees a reader, sure, get fancy. For something that will live in pockets and terminals, I don’t gamble with corner geometry.
A quick standards reality check (ISO vs ANSI)
People say “ANSI credit card” like it’s a totally separate thing. Functionally, most of the industry converges on the same familiar footprint, but the language differs, and that’s where mistakes creep in.
ISO/IEC 7810 ID‑1
The international baseline: 85.60 × 53.98 mm, rounded corners, defined thickness class, and cross-compatibility expectations.
“CR80” / common US vendor shorthand
Often quoted as 3.375 × 2.125 in. That 3.375 in value is slightly different from a direct mm-to-inch conversion of the ISO number. Vendors may still produce within acceptable tolerance for devices, but don’t assume, confirm what they’re cutting to.
A specific reference, since people like receipts for numbers: ISO/IEC 7810 is the standard that defines ID‑1 dimensions (ISO/IEC 7810:2019). If you need chapter-and-verse tolerances, you’ll want the actual document from ISO.
Printing, branding, and the lie of “centered”
Now, this won’t apply to everyone, but if your design has tight borders or “perfectly centered” frames, you’re asking for pain.
Cutting variation + print registration variation + lamination creep = your perfect border becomes visibly uneven. Human eyes are brutal about symmetry. They’ll ignore a slightly off logo; they won’t ignore a 0.3 mm uneven frame around the edge.
Better strategy (in my experience): design like you expect drift. Give the layout room to breathe. Use backgrounds that tolerate slight movement. Keep microtext and thin rules away from edges unless you enjoy reprints.
Thickness: not just a number, a feel
0.76 mm sounds like a spec. It’s also a tactile promise.
Too thin and the card feels like a flimsy promo badge. Too thick and you can run into reader issues, wallet bulge complaints, or lamination failures (depending on materials). Add heavy coatings, metallic layers, or tactile finishes and you may accidentally “grow” the card past what devices tolerate.
Look, people will forgive a weird hologram. They won’t forgive a card that doesn’t tap consistently or sticks in an ATM.
Manufacturing notes that save real projects
Some of this reads boring until you’re the one explaining delays.
– Edge finishing: bevels/chamfers change perceived size and can affect insertion feel
– Flatness: warped cards cause intermittent chip contact and feeding errors
– Roughness & coatings: too much texture can increase friction in feeders (and scuff faster)
– Tool wear: corner radii drift as punches age; track it or it will track you
And if you’re embedding a chip module, alignment isn’t “close enough.” It’s a hard requirement tied to terminal contact geometry and module placement conventions.
Testing: the unglamorous part that prevents field failures
I like prototypes. I trust validation more.
A practical validation plan usually includes:
– dimensional checks (width/height/thickness across samples)
– corner radius verification against a gauge
– flatness/warp checks after lamination and after environmental exposure
– abrasion testing for print durability (especially on edges)
– reader trials: swipe, insert, tap, repeat, then repeat again after wear
Record drift by lot. Track what changed. Tooling, substrate batches, cure times, lamination pressure, one of them will be the culprit when things go sideways.
Wallet apps and “digital cards” (still relevant to dimensions, oddly)
Digital wallets don’t have physical tolerances, but the UI still borrows the ID‑1 proportions because users recognize it instantly. If you’re designing both physical and digital experiences, keeping consistent aspect ratios and safe zones reduces cognitive friction. People shouldn’t have to “re-learn” your card in Apple Pay just because your plastic had different constraints.
Different medium, same mental model.
The take nobody loves: standards win
You can fight the ecosystem, but you won’t win.
Build to ISO ID‑1 geometry, respect corner radii, treat thickness as a system constraint, and design print layouts that forgive manufacturing reality. Do that and your card will behave like a card, quietly, reliably, everywhere it needs to.