EPSS vs. CVSS: Complementary, Not Competing

Why exploitation likelihood (EPSS) and severity (CVSS) answer different questions, and prioritization needs both.

Why It Matters

Using CVSS alone to prioritize a backlog systematically over-invests in high-severity-but-unlikely findings and under-invests in moderate-severity findings under active, real-world exploitation.

Application to Application Security

A dependency CVE with a high CVSS score but near-zero EPSS is a very different priority than one with a moderate CVSS score and a high, rising EPSS — the second is the one attackers are actually using right now.

Relation to Software Development

This is precisely the argument Strata makes in its own vulnerability-prioritization guidance: severity alone doesn't determine what to fix first, and combining it with exploitability and business context does.

How Strata Surfaces This

Strata does not calculate CVSS or EPSS scores for its findings; each finding carries a severity rating. Combining severity with EPSS and business context is the subject of Strata's dedicated prioritization article, linked below.

Developer guidance

Build a triage rule that combines both scores — e.g., high CVSS + high EPSS gets fixed this sprint, high CVSS + low EPSS gets scheduled, not dropped.

Management guidance

A prioritization framework that visibly combines severity and exploitability is easier to defend to stakeholders than "we fixed things in CVSS order," which invites the obvious follow-up: why wasn't the highest score first?

FIRSTEPSSSource ↗