Fair credit is the awkward middle of the market. A FICO score in roughly the 580 to 669 range is too high to need a secured card and too low for the cards that advertise flat 2% back or large travel bonuses. The useful question is not which card pays most, it is which rewards you can get at this tier without paying more in fees and interest than the rewards are worth. For most people at a fair score, that calculation does not come out the way the marketing suggests.
The winner
At a fair score the winner is the Capital One Platinum ($0). No annual fee, no rewards to distract you, and an automatic review for a higher limit. At this tier the goal is to build history at zero cost and cross into good credit, not to chase rewards.
When to pick something else. If you will spend more than about $2,600 a year and pay in full, the QuicksilverOne ($39) earns its fee back at 1.5%. If you might carry a balance, a credit-union card with a lower APR saves you more than any rewards rate.
What to avoid. Treat the 5% and 10% rates advertised at this tier as a warning sign, not an attraction, and avoid anything with a program fee or monthly servicing fee. At fair-score APRs the interest matters far more than the rewards.
What a fair score actually unlocks
Lenders generally treat 670 as the start of “good” credit. Below it, approvals cluster in a specific part of the market:
- Usually available: Capital One’s rebuilding tier, credit union cards, some Discover products, store cards, and secured cards if an unsecured application is declined.
- Usually not available: Chase’s Sapphire and Freedom cards, the Amex rewards line-up, and the mainstream flat 2% cards, which typically ask for good to excellent credit.
There is no public score cut-off at any issuer, and approval depends on income, existing debt, recent applications and the rest of your file as well as the score. Two people with the same number get different answers.
The cards worth considering
Capital One QuicksilverOne: $39 annual fee. 1.5% back on everything with no caps, plus 5% on hotels and rental cars booked through Capital One Travel. The published variable APR has sat around 29%. Accounts are reviewed automatically for a higher credit line in as little as six months.
The fee math is straightforward: $39 divided by 1.5% is $2,600 of annual spending to break even. Above that you are ahead; below it you are paying for the privilege of earning. For someone putting $400 a month through the card, it returns about $72 a year against the $39 fee, a real but modest $33.
| Yearly spending on the card | QuicksilverOne, after the $39 fee |
|---|---|
| $1,200 | −$21 |
| $2,600 | $0 |
| $5,000 | +$36 |
| $10,000 | +$111 |
Capital One Platinum: $0 annual fee, no rewards at all, published variable APR around 29%. It looks unattractive next to the card above, and for a specific purpose it is the better choice: it costs nothing to hold, it builds history, and Capital One has a well-established route to product change it into a rewards card once your file improves. Paying $39 a year for 1.5% while rebuilding is optional; building history for free is not.
Capital One Platinum Secured: $0 annual fee, with a refundable deposit of $49, $99 or $200 for an initial $200 line, the tier set by the issuer rather than chosen by you. If an unsecured application is declined, this is the cheapest way in, and the $49 tier is the lowest entry cost anywhere in this market.
Discover’s cash back cards have historically approved applicants across a wide score band and pay 5% on rotating quarterly categories on up to $1,500 of spending per quarter, with activation required, plus a Cashback Match that doubles everything earned in the first year. That makes a Discover card the highest-paying option at this tier in year one and an ordinary 1% card afterward. One caveat: Discover is being integrated into Capital One, and the availability of individual Discover products has changed more than once during 2026, so confirm what is actually open to new applicants before you plan around it.
Credit union cards deserve more attention than they get at this tier, for one reason that has nothing to do with rewards. Federal credit unions operate under an interest-rate ceiling set by their regulator, 18%, which the NCUA extended in February 2026 through September 10, 2027, which is far below the high-twenties rates typical of cards marketed to fair credit. If there is any chance you will carry a balance, that difference is worth more than every rewards program on this page combined. Membership usually requires meeting a field-of-membership rule, which is often as simple as living in a particular area or joining an affiliated association.
How we judged these cards: on what they cost to hold and to borrow on first, and on what they pay back second, using each issuer’s published fees and rates as of October 2026. Approval is described as a pattern, never a promise, because no issuer publishes a score cut-off. No issuer pays CardRankr, and we earn nothing if you apply.
What to avoid, with the arithmetic
This is the part of the market where the worst products live, and they are marketed hardest at exactly the people reading this page.
Stacked fees. Some cards aimed at fair credit charge an annual fee and then a separate “premium card” or account-servicing fee on top, for example $95 a year plus roughly $30 for the physical card, on a product with an APR near 30%. That is $125 of fixed cost. At 5% back on dining and gas you would need $2,500 a year in those two categories alone just to break even, on a card whose opening credit line may be $300 to $500. The structure does not work, and the generous-sounding rates are what make it look as though it might.
Monthly maintenance charges. Any card charging a fee every month is best avoided outright. A $9 monthly fee is $108 a year, which no rewards rate at this credit tier can cover.
Deferred-interest store promotions. “No interest if paid in full in 12 months” is not the same as 0% APR. If any balance remains at the end of the promotional period, interest is typically charged retroactively on the entire original purchase from the date of sale. A single month’s slip can turn a $1,200 sofa into a $1,500 one. The CFPB’s explanation of these offers is worth two minutes before you sign.
At this tier, the APR matters more than the rewards
This is the single most important point on the page, so here is the comparison in full.
A card with a 29% APR carrying a $1,000 balance for a year costs roughly $290 in interest. The same card paying 1.5% on $5,000 of annual spending earns $75. The interest is close to four times the rewards.
The conclusion follows directly: if you carry any balance at all, choose the lowest APR you can get and ignore rewards entirely. Rewards cards are for people who pay in full every month. At a fair score, with rates near 30%, that distinction is not a nuance, it is the whole decision.
Moving from fair to good
Five things do the work, in rough order of impact:
- Never miss a payment. Payment history is about 35% of a FICO score, and a single 30-day late mark can stay on your report for years. Autopay for at least the minimum, set the day after payday.
- Keep reported utilization low. Utilization is about 30% of the score and is calculated from your statement balance, not from whether you pay in full afterward. On a $500 limit, a $150 statement balance reports 30%; paying down before the statement closes reports less. Under 10% is where the scoring benefit is largest.
- Ask for credit line increases every six months. A higher limit lowers utilization at the same spending level, the one score factor you can improve without waiting. Capital One reviews automatically; other issuers require asking, and some do so without a hard inquiry.
- Check your reports for errors. You can access your reports from all three bureaus free at AnnualCreditReport.com, and disputing an inaccurate late payment or an account that is not yours is the fastest score change available to anyone.
- Do not close your oldest card. Length of history is about 15% of the score, and closing an old account also removes its limit from your utilization calculation.
A realistic timeline from fair to good, with no negative marks being added, is six to eighteen months. Nothing legitimate is faster, and anyone promising faster is selling something.
The product-change route, which saves you an application
When your file improves, the instinct is to apply for a better card. Often the better move is to ask your existing issuer for a product change: converting the account you already hold into a different card in the same family.
A product change keeps the account’s age, usually involves no hard inquiry, and does not add a new account to your file. Capital One’s Platinum to Quicksilver path is the best-known example at this tier. Use the words “product change” or “upgrade” rather than asking to apply, and ask explicitly whether the account number and opening date are preserved.
Once you are reliably above 670, the mainstream market opens up: see our no-annual-fee picks and cash back round-up. For the mechanics of building from a thin file, see building credit, and if an unsecured application is declined, secured cards. First card of any kind? Start with cards for beginners.
The bottom line at a fair score
At a fair score the honest ranking is: a credit union card if you might carry a balance, the no-fee Capital One Platinum if you want to build history at zero cost, and QuicksilverOne if you will spend more than about $2,600 a year and pay in full. Avoid anything with stacked or monthly fees, and treat the 5% and 10% rates advertised at this tier as a warning sign rather than an attraction. The goal is not to maximize rewards for the next twelve months, it is to cross 670 and stop shopping in this part of the market.
Fees, APRs, approval criteria and product availability all change without notice, and approval always depends on the issuer’s own assessment of your full file. Confirm current terms before you apply.
At the APRs typical of this tier, the interest matters far more than the rewards. How credit card interest actually works sets out the arithmetic, and authorized user vs joint account vs co-signer covers the fastest way to add good history to a thin file.