Join our Newsletter — 33% off our NHI Course

NIST SP 800-90

NIST SP 800-90 is a family of standards for generating random numbers and deterministic random bit sequences for cryptographic use. It covers methods such as entropy sources, random bit generators and deterministic random bit generators, which are used to produce keys, nonces and other secrets with defined security properties.

NIST SP 800-90 as a randomness standard family

NIST SP 800-90 is not a single algorithm. It is a standards family for producing cryptographically suitable random bits through entropy sources, random bit generators, and deterministic random bit generators, so security systems can create secrets with defined properties.

Its practical value is that it turns “random enough” into a controlled design requirement. Key generation, nonces, salts, session material, and other security tokens depend on a generator that is both unpredictable and correctly instantiated, because weak randomness can undermine an otherwise strong system.

What entropy and deterministic random bit generators do

The family separates the raw unpredictability problem from the output problem. Entropy sources measure and accumulate uncertainty, while deterministic random bit generators expand a seed into usable output without becoming predictable once the seed is known or compromised.

This distinction matters because many failures come from confusing source quality with output quality. A well-designed DRBG can still fail if it is seeded poorly, reseeded incorrectly, or used in a way that ignores its security strength and instantiation rules.

Why random-number quality is a security boundary

Randomness is part of the trust base for encryption keys, authentication tokens, API keys, nonces, and many protocol protections. If an attacker can predict output, they may be able to impersonate systems, replay messages, or recover protected material even when the surrounding cryptography is sound.

Entropy failures are especially dangerous in early boot, virtualised, embedded, or poorly initialised environments where the system may not yet have enough uncertainty to seed a generator well. In those cases, the weakness is often invisible until secrets are reused or an external observer notices a pattern.

For practical reading on this topic, NHIMG’s Random Number Generation and Entropy in Security Guide connects SP 800-90 to real implementation failure modes, including token length, boot entropy, and generator misuse.

How practitioners should interpret and apply the family

NIST SP 800-90 is best treated as a design and assurance reference, not a checkbox. Teams should understand which component is supplying entropy, which component is expanding it, and how the generator behaves across restarts, scaling events, and platform changes.

That often means validating the full chain, from platform entropy availability through the library or OS interface that applications actually call. If the application depends on language defaults, container images, or virtual machine startup state, the real control point may sit lower than the code that consumes the random bytes.

Practitioners should also treat output reuse and long-lived secrets as a design smell. When a random source is central to identity material, encryption material, or protocol freshness, weaknesses tend to surface as large-scale compromise rather than isolated bugs.

Standards & Framework Alignment

This section maps relevant standards and security frameworks to the operational risks and controls described in this guidance.

NIST SP 800-57 and NIST SP 800-53 Rev 5 set the governance and control requirements practitioners need to meet.

Framework Control / Reference Relevance
NIST SP 800-57 Key Management SP 800-90 generates key material that key management must protect and refresh.
Recommendation — Treat entropy quality as a prerequisite for strong key lifecycle decisions.
NIST SP 800-53 Rev 5 IA-5 — Authenticator Management Random generation underpins authenticators, tokens, and other secret material used in access control.
SC-13 — Cryptographic Protection The standard family supports cryptographic operations by producing secure random values and seeds.
SI-10 — Information Input Validation Poor entropy input or generator seeding is a technical failure mode that must be constrained and checked.
Recommendation — Use IA-5 to manage token and secret generation with trustworthy randomness. Apply SC-13 to ensure cryptographic processes rely on approved random generation. Validate entropy inputs and generator configuration before relying on produced secrets.