Exploit Prediction Scoring System
The Exploit Prediction Scoring System (EPSS) is a scoring system that estimates the probability that a specific software vulnerability will actually be exploited by attackers. Rather than describing how severe a flaw is, it focuses on how likely it is to be used in the real world. Security teams often use these probability scores to help prioritize which vulnerabilities to fix first.
EPSS is a data-driven, machine learning-based framework that estimates the likelihood of a given Common Vulnerabilities and Exposures (CVE) entry being exploited in the wild, typically expressed as a probability over a 30-day forward-looking window. The output is a probability score between 0 and 1 that reflects predicted exploitation activity, which is distinct from severity-oriented scoring systems such as CVSS. In practice, organizations combine EPSS probabilities with severity and business-context data to prioritize remediation, though the value of the score depends on the quality of the underlying model and how it is integrated into an organization's vulnerability management process.
Why it matters
Most organizations face far more known vulnerabilities than they can realistically remediate at any given time. Traditional severity-based scoring, such as CVSS, describes how damaging a flaw could be if exploited, but it does not indicate how likely that exploitation actually is. This gap often leads teams to spend effort patching high-severity vulnerabilities that attackers are unlikely to target while overlooking lower-severity flaws that are being actively weaponized. EPSS addresses this by estimating the probability that a specific vulnerability will be exploited, giving security teams a signal about real-world attacker behavior rather than theoretical impact alone.
Who it's relevant to
Inside EPSS
Common questions
Answers to the questions practitioners most commonly ask about EPSS.