A card sits in a drawer for a year. You finally reach for it — boarding a flight, checking into a hotel, paying for car repairs — and the terminal shows decline. You log into the app on your phone; the account is open, the balance is zero, the credit limit hasn't moved. The bank hasn't closed the card. It has done something quieter and more inconvenient: placed it on a dormancy verification hold.
The hold is the step before closure, and very few cardholders know it exists until they're standing at a register watching a $400 transaction get rejected on a card their app still lists as "active."
The middle state: open but unusable
Most coverage of inactive credit cards describes two states. The card is open and works, or the card is closed and doesn't. Inside the issuer's authorisation system there is a third: the account is flagged for prolonged dormancy and the next charge attempt is blocked at the authorisation layer, pending a human verification call. The card is open. The limit is preserved. The tradeline reports normally to the bureaus. But the next transaction will not authorise.
This is a soft block, not a closure. There is no remark code on your credit report and the score impact is zero. Closure remark codes are not in play. The only thing that has changed is a flag inside the issuer's fraud-and-dormancy overlay — one that any new charge will hit before it ever reaches your account balance.
What triggers the hold (versus a clean approval)
The hold fires when the issuer's risk model sees a transaction it considers improbable given the recent transaction history. Two ingredients almost always combine:
- A long dormancy gap. Typically 9–18 months since the last posted charge, depending on the issuer. The exact threshold for the hold is shorter than the threshold for closure — Capital One's closure window is 24 months, but the verification hold fires at around 12.
- A transaction the model considers high-risk relative to the account's idle state. A high-ticket purchase ($300+), a foreign merchant, a card-not-present transaction at a new merchant, or a card-present swipe far from the cardholder's billing zip.
The model treats this combination the way it treats any other potential fraud event: as something to verify before it lets through. A dormant card surfacing for a four-figure purchase looks statistically a lot like a card a thief found in a drawer. The fact that the cardholder is the thief in their own pocket is not something the model can distinguish.
A small posted charge — $5 at a familiar merchant, in your home zip — almost never triggers the hold. This is one of the reasons a recurring tiny charge works: it both clears the inactivity timer and trains the fraud overlay that this card is in normal use.
Per-issuer behaviour: who does this, and what they call it
The mechanic exists at every major US issuer; the customer-facing label varies.
American Express. The most aggressive verification regime. Amex internally calls long-dormant accounts "decision-suspended" and routes the next attempted authorisation to a fraud-protection queue. The decline reason at the terminal usually reads "call issuer" or "do not honour." The reactivation call takes 5–10 minutes and includes one identity question and one transaction-confirmation question. Amex is also the issuer most likely to leave the hold in place if the cardholder doesn't call — the account quietly stays unusable until the broader inactivity window catches up.
Capital One. Holds usually fire at the 12–18-month mark on consumer cards. Capital One's terminal decline reason is typically generic ("transaction declined") but the issuer texts the cardholder within a minute of the decline with a prompt to confirm or deny. Confirming via the text usually clears the hold without a phone call — the most customer-friendly version of the mechanic.
Chase. Fires the hold less aggressively than Amex or Capital One, but when it does, it tends to require a full phone call and a security-question identity check. The decline reason often reads "contact card issuer." Chase will sometimes lift the hold and immediately re-run the authorisation while you are still on the phone.
Citi and Bank of America. Both move from "open" to "closed" faster than most issuers (the 6-month mark is realistic for closure), which leaves less room for the verification-hold middle state. When holds do fire, they tend to be brief and clear with a single transaction-confirmation prompt in-app. Full per-issuer timeline breakdown.
Discover. The least likely to trigger a verification hold. Discover's preference is to send written warnings and then close, rather than block authorisations and force a call. If you see a Discover card decline on dormancy, the closure is usually already in motion.
What happens at the merchant terminal
The decline reads as a generic "do not honour" or "call issuer" on the merchant's receipt. The merchant has no idea this is a dormancy event — to them, it's a normal authorisation rejection, indistinguishable from a fraud-blocked transaction or an over-the-limit attempt. They will ask you to try a different card. If you try the same card again on the same terminal in the next hour, it will decline again, because the hold is on the account not the transaction.
The hold timing is unfortunate by design. It tends to fire on the first meaningful transaction after a long gap, which is exactly when cardholders reach for the card they haven't used in a while. The classic profile: a travel card pulled out at airport check-in after twelve months in a drawer, declining the bag-fee charge in front of a queue. A backup card brought out for an emergency repair, declining the parts charge while the vehicle is on a lift.
The verification call: what they actually ask
The call is shorter than the average customer-service call but has a specific structure. You will be asked three things:
- Confirm identity. Last four of your social, the address on file, the email on file, sometimes the date of the last payment. Typical security-questions stack.
- Confirm the specific declined transaction. "We saw an attempted charge of $437.20 at [Merchant] in [City]. Was that you?" If you confirm, the hold drops in real time. If you deny, the card is treated as compromised and routed to a fraud queue with a card reissue.
- Confirm intent to keep the account open. Asked rarely, but Amex in particular sometimes uses the call as an opportunity to ask whether you want to keep the card. The right answer if you do is unambiguous yes. A lukewarm answer can put the account on a closure track for the next review cycle.
Once you complete the verification, the rep will tell you the card is unblocked. Run the same transaction again within five minutes and it should authorise cleanly. The hold does not require any follow-up activity on your end to stay cleared.
Why this is not the same as a closure
The verification hold is mechanically distinct from an inactivity closure in four ways. The credit limit is preserved — closures remove the limit from your total available credit, which raises utilization across your portfolio. The account stays open on your credit report — a closure adds a "closed by credit grantor" remark code that stays visible for ten years. The reactivation path is a phone call, not a new application — closure reactivation is much harder, especially after 90 days. And the rewards balance is intact — closure can forfeit points, while the hold doesn't touch them.
The cost of a verification hold is friction, not financial. The cost of the closure it precedes is real.
Cards most likely to land in verification limbo
The hold-then-closure pattern shows up most often on three card profiles. Travel and rewards cards used only abroad — Sapphire Reserve, Venture X, Amex Platinum — sit dormant between trips and decline the first foreign-merchant charge after a long gap. Backup or emergency cards kept for medical or auto-repair use sit unused for years and decline the high-ticket charge they were kept for. Premium store cards opened for a single big-ticket purchase (an appliance, a furniture set) decline the next discretionary use.
The common thread is that the next use of these cards is exactly the kind of transaction the fraud overlay is sensitive to — large, urgent, or unusual relative to baseline. The cardholder's pattern is "I'll use it when I need it," and the issuer's model is "this account hasn't been used and is suddenly making a high-risk charge."
The cheap fix
The fix for the verification hold is the same fix that prevents the closure that follows. Post one small charge inside the issuer's review window, on each card you want to keep available. The hold's risk model reads recent activity as a baseline and a tiny monthly charge resets that baseline without much else. A $1–$5 posted charge per month is enough to keep both the hold and the closure off the account.
You can run that charge manually (a quarterly $1 digital purchase), anchor it on a subscription you would buy anyway, or let an automation tool handle it across multiple cards. The volume of the charge does not matter — the inactivity review and the fraud overlay both check the date, not the amount. What matters is that something has posted recently enough to make the next high-ticket use look like part of an active pattern, not a return from the dead.