Authentication Factors and Methods

Key Takeaways

  • A factor is knowledge, possession, or a biometric characteristic. Count independent types of evidence rather than prompts.

  • A password and a manually entered one-time code can use two factors without providing phishing resistance.

  • Use verifier-bound cryptographic authentication where the requirement is phishing resistance.

  • Test enrollment, login and recovery as separate paths for the chosen method.

Last updated: November 2026

Identify the factor behind each check

An authenticator is something the claimant uses to prove control of identity credentials. NIST SP 800-63B-4 describes knowledge, possession and inherence. A password is knowledge. Control of a physical or cryptographic authenticator can provide possession. A biometric match can provide inherence. In this NIST model, biometrics are used with a physical authenticator, and the biometric characteristic by itself is not an authenticator.

Multifactor authentication (MFA) uses more than one distinct factor. A password and an independently verified phone authenticator can provide knowledge and possession. A password and another memorized secret are both knowledge even if the screen presents them as two separate steps. A device’s local personal identification number (PIN) can activate a cryptographic authenticator. In that arrangement the PIN provides knowledge and the cryptographic key supplies possession when the required checks are enforced. A single device can support multiple factors. For example, a security key that produces a cryptographic response only after checking a local PIN can combine key possession with knowledge. A key that merely requires a touch to indicate user intent does not thereby add a second factor.

MethodWhat the verifier or device checksKey selection consideration
PasswordThe claimant presents the memorized secretKnowledge. A phishing site can collect it
Password and a manually entered one-time passwordKnowledge plus control of the code-producing authenticatorThe user can still relay the short-lived code to an impostor
Password and a text message codeKnowledge plus control of an out-of-band channelNIST treats public telephone network use for out-of-band authentication as restricted
Security key using Web Authentication (WebAuthn) with enforced user verificationA signed challenge bound to the relying party, with a local activation factorEvaluate factor enforcement and key lifecycle for the required risk level
Password and a second memorized PIN checked centrallyTwo knowledge promptsThe second prompt does not supply a different factor

Work a phishing-resistance requirement

Suppose a helpdesk operator uses a password plus a time-based one-time password (TOTP). Their organization now requires evidence that the selected method prevents a fake sign-in page from collecting a response it can use at the genuine verifier. A fraudulent page can ask the operator to enter their valid short-lived code and relay it promptly. A short validity window and one-time use help against some replay scenarios. The code is still not cryptographically bound to the genuine verifier’s session merely because it expires.

A properly implemented WebAuthn authentication can instead bind the cryptographic proof to the relying party identifier. With an approved security key and the required user verification settings, the authenticator validates the expected user interaction and signs a challenge for the relevant site. A response produced for an impostor’s relying party identifier does not satisfy the intended relying party’s binding check. Verify the actual method and configuration. The brand of an authenticator or the presence of two prompts does not establish phishing resistance by itself.

Verify enrollment and fallback

Test registration of a second allowed device if the policy permits one. Verify the normal sign-in event, a failed response, and the supported lost-device recovery process. For a code-based method, test an expired code and a repeated use of a code the verifier already accepted. For a security key, check that the required local activation factor is enforced, including on a second enrolled device. If an administrator can bypass a stronger sign-in path with a weak recovery path, the effective protection depends on how that path is governed. During the test, record which action required which evidence and whether the fallback followed the approved policy. Control the test environment so an operator does not lose access while evaluating recovery.

Factor selection and an assurance label are related. A two-step sign-in is insufficient to claim a particular NIST assurance level. Confirm which factors are enforced, the cryptographic and phishing-resistance properties, and the other applicable requirements in the current standard. An external identity provider can perform these checks during federation. The relying party still validates the delivered assertion, and application authorization still determines what the resulting session can access.

See NIST’s SP 800-63B-4 authenticator and verifier requirements, including its verification and phishing-resistance requirements.

Test Your Knowledge

A sign-in requires a memorized password and a different centrally checked memorized PIN. Which factor types does this require?

A

Possession and inherence

B

One factor type, knowledge, used in both checks

C

Knowledge and possession because two prompts appear

D

Possession only because a web browser displays both prompts

Test Your Knowledge

An administrative login requires phishing resistance. Which choice fits the documented requirement when properly implemented?

A

A password and a manually typed time-based one-time code

B

A password and a code read out from a text message

C

A WebAuthn authenticator whose proof is bound to the relying party

D

Two different passwords memorized by the same administrator

Sections you finish are checked off in the contents.