How We Rate Credit Cards

Every credit card rating on this site is generated using a category-specific scoring model. Each model uses defined factors, weights, and scoring rules based on the characteristics that matter most for that type of card. The same methodology applies whether or not we have a commercial relationship with the issuer.

261 cards rated · 11 categories · 31 data points per card · 6–7 weighted factors per model

The short version

We rate cards on a 5-point scale in tenths of a star.

A cash back card and a balance transfer card are judged by different rules because they serve different purposes. Within a category, a higher score means a stronger card for the typical person shopping for that type of card. Scores across different categories are not directly comparable.

Each rating is calculated using the model assigned to the card's primary category. We define the factors, assign their weights, establish the scoring rules, and apply the same methodology to every card in that category.

The result is a rating produced by the model rather than a score assigned based on an editor's overall impression.

How the models work

Each card enters our scoring system as 31 separate data points — including fees, introductory periods, earning rates, caps, credits, protections, and eligibility requirements.

Each category model uses the specific data points that matter for that type of card and combines them into six or seven weighted factors.

Every factor has a defined scoring rule. Depending on the category, that might mean mapping the length of an introductory APR period, a balance transfer fee, a rewards rate, an annual fee, or another product characteristic to a score.

Each model emphasizes the factors that matter most for the job that type of card is intended to do. A long 0% period, for example, matters much more to someone transferring debt than to someone choosing a premium travel card.

Where a model requires assumptions about spending or card usage, we use a consistent benchmark within that category rather than changing the assumptions from card to card.

Below are examples of what some of our category models emphasize.

Balance transfer cards

The length of the balance transfer introductory APR period carries the most weight because additional interest-free time is usually the main reason to choose one of these cards.

The balance transfer fee is also an important part of the score. The difference between a 3% and 5% fee on a $6,000 transfer is $120, so a longer introductory period does not automatically make one offer better than another.

The model also considers factors such as how long you have to complete a qualifying transfer, the card's availability, and the value it provides after the introductory period ends.

Cash back cards

The rewards a typical cardholder can realistically earn carry the most weight.

We use a consistent spending profile rather than assuming cardholders maximize every bonus category. A high rewards rate on a major spending category therefore contributes more value than the same rate on a relatively small spending category.

The model also considers the welcome offer, earning caps, activation requirements and other earning restrictions, redemption rules, additional benefits, availability, introductory APR offers, and annual fee.

Travel cards

Earning power and the value of the rewards earned are major components of the score. That includes factors such as redemption value and, where applicable, access to airline and hotel transfer partners.

We also weigh the annual fee against the realistic value available from the card's benefits and credits. A high annual fee can still score well when the card provides enough usable value to offset it.

Other factors include travel protections and benefits, the welcome offer, introductory APR features, and access to the card and its rewards ecosystem.

Business cards

Rewards fit for a typical small business carries significant weight. Depending on the card, relevant spending may include office supplies, internet and phone service, shipping, advertising, gas, dining, and general business purchases.

The model also considers the welcome offer, annual-fee-to-value tradeoff, rewards ecosystem, benefits, and the structure of the product.

For example, a product that requires the balance to be paid in full can serve a different purpose than a traditional credit card that allows balances to revolve. The model accounts for differences like these when they are relevant to the category.

Rewards cards

Earning power and redemption value carry substantial weight, along with how practical it is for a typical applicant to obtain and use that value.

We also account for requirements that can make a card's advertised rewards harder to earn, such as paid memberships, minimum account balances, banking relationships, invitation-only eligibility, and limits on how much spending earns the higher rate.

A headline rewards rate carries less weight when significant requirements stand between the cardholder and that rate.

What the ratings mean

4.8–5.0 — Category leader

Among the strongest cards in the category for the typical person shopping for that type of card. Even top-rated cards have tradeoffs, which we explain in our reviews.

4.5–4.7 — Excellent

Very strong cards within their category, often particularly well suited to certain types of cardholders.

4.0–4.4 — Solid, with tradeoffs

Good products whose value depends more heavily on how you plan to use them. Their benefits can be meaningful, but so can their limitations.

Below 4.0 — More limited value

These cards may underperform others in their category, carry costs or restrictions that reduce their value, or make the most sense only in narrower circumstances.

The lowest ratings generally go to products whose costs, restrictions, or limited value make them difficult to recommend for most people.

How we apply the methodology

Each card has one primary rating

Each card receives a rating based on the category that best represents what the card is primarily designed to do.

A cash back card that also includes a balance transfer offer, for example, may still be evaluated primarily as a cash back card. If introductory financing is the defining feature of the product, it may instead be evaluated under the applicable introductory APR or balance transfer methodology.

That primary rating is the rating used across the site.

We rate cards based on realistic use

Our models are designed around what a cardholder can reasonably expect to receive rather than the maximum theoretical value that could be extracted from a card.

Benefits or earning rates that require substantial assets, paid subscriptions, unusual spending patterns, invitation-only access, or other significant requirements are accounted for accordingly.

When a card is specifically designed for a particular group, such as renters or small business owners, we evaluate it from the perspective of the people the product is intended to serve.

We use standard public offers

Ratings are based on the standard public offer generally available to qualifying applicants.

Targeted offers, personalized offers, referral bonuses, and other offers that are not broadly available do not increase a card's rating.

When the public terms of a product materially change, the updated terms can change its score.

The model determines the rating

Cards within the same category are evaluated using the same factors, weights, and scoring rules.

We do not manually raise or lower a rating because we have a commercial relationship with an issuer — or because we do not.

If a card we do not work with earns a higher score than a card we do work with, it keeps the higher rating.

The scale is designed to distinguish between cards

Our ratings are not designed to place nearly every product near the top of the scale.

The models are intended to distinguish between exceptional products, strong products, more specialized products, and products whose costs or limitations make them less competitive within their category.

How compensation affects our ratings

We may receive compensation from some partners when you click our links and take certain actions on their websites.

Compensation is one of several factors that may affect which cards we feature and where they appear on a page. For more information about how our advertising relationships work, see our Advertiser Disclosure.

Compensation does not affect a card's rating.

Partner and non-partner cards are evaluated using the same scoring models and scoring rules. We do not change a model's weights or scoring criteria based on whether we have a commercial relationship with an issuer.

A partner cannot pay for a higher rating or award.