Card testing is a type of credit card fraud where criminals use a real website's payment form to check whether stolen card numbers still work. They are not trying to buy anything from you. They want a yes or no from the bank, and your checkout gives them one.
It goes by other names too: carding, card checking, BIN attacks and enumeration attacks. They all describe the same thing: many cards, small amounts, very fast.
Where the cards come from
Stolen card numbers are sold in bulk, collected from data breaches, phishing pages and skimmers. A list is cheap, and much of it is dead: cards that were cancelled, expired or already reported. Before a buyer puts any effort into a list, they want to know which cards are live, because a working card sells for far more than an untested one.
How a card testing attack works
- The fraudster finds a site with a low minimum payment and little screening: a small store, a donation form, a subscription with a $1 trial.
- A script submits card after card, often hundreds an hour, rotating IP addresses through proxies or rented servers so each attempt looks like a new visitor.
- Each approval or decline tells them which cards are good. Approved cards go on a "checked" list and get used or resold for bigger purchases somewhere else.
A variation, the BIN attack, starts without stolen numbers at all. The attacker takes the first six or eight digits of a real bank's card range (the BIN), generates the rest so they pass the basic checksum, and guesses expiry dates and security codes until something goes through.
Why your site gets picked
Testers look for the path of least resistance: no rate limit on the payment step, small amounts allowed, error messages that say exactly why a card failed, and a checkout that works without an account. Donation pages and cheap digital goods are favourites, because there is nothing to ship.
What it costs you
- Fees on every authorisation attempt, approved or not.
- Chargebacks weeks later, when the real cardholders spot the test charges.
- Trouble with your payment provider. A sudden flood of declines can trigger a review, higher fees or a frozen account. Card networks now watch for testing at the acquirer level too, so your provider has its own reasons to act fast.
- Polluted data: fake orders, fake customer records and skewed conversion numbers.
How to spot an attack
The pattern is hard to miss once you look. The tell-tale signs:
- A sudden burst of payment attempts, often for the smallest amount you sell.
- A decline rate far above normal: most of the cards fail.
- Many different cards, but the same few IP networks behind them.
- Cards from many different banks and countries, sometimes with BINs in sequence.
- Attempts at odd hours, at a steady machine-like pace.
It tends to spike in the weeks before Black Friday, because fraudsters want a list of working cards ready for when big orders blend in with everyone else's.
How to shut it down
Check the connection first. Card testing runs from rented servers and proxies almost every time, because the testers need to rotate addresses cheaply. Checking the IP before you send the payment to your processor stops most of it at the door, and those attempts never cost a fee.
Count cards per network. Real customers try one card, maybe two. If one network sends ten different cards in five minutes, block that network for the rest of the day.
Check the BIN against the IP. A card from one country used from a datacenter in another is a classic testing pattern. Neither signal alone proves anything, but together they are strong.
Raise the floor. If you have a product or donation amount under a dollar or two, that is what testers will find. Consider a higher minimum.
If you are under attack right now, blocking payment attempts from hosting and proxy networks at checkout is the fastest single change; then add a per-network card limit. For the full list, see our card testing prevention checklist.
FAQ
Does card testing hurt me even when the payments fail? Yes. Every declined attempt can cost a processing fee, a high decline rate can get your payment account reviewed or frozen, and the cards that do go through become chargebacks weeks later.
Why check the visitor's network? Most testing comes from datacenter servers, proxies and VPNs rather than home connections, so checking the network before the payment call blocks the bulk of it without touching real buyers.
What is a BIN attack? A variant that needs no stolen numbers: the attacker takes a real bank's BIN range, generates card numbers that pass the checksum, and guesses expiry and security codes until one is approved.
The short version
Card testing is cheap for fraudsters and expensive for you. Most of it comes from datacenter servers, proxies and VPNs rather than home connections, which is why checking the visitor's network before the payment call is one of the most effective defences. Fraudex does that check, and the card's BIN lookup, from one API key.