CVSS: Common Vulnerability Scoring System
CVSS (Common Vulnerability Scoring System) is an open standard, maintained by FIRST, for rating how severe a vulnerability is on a 0 to 10 scale. It works by scoring a small set of metrics, how the vulnerability is reached and exploited, and what it impacts, and combining them into a Base score, with optional Temporal/Threat and Environmental scores layered on top for real-world context. Four versions are in active use today: v2.0, v3.0, v3.1, and v4.0, each with different metrics and formulas.
Score a vulnerability right now
Free interactive calculator, all four versions, plain-language help for every metric.
How Base scoring works
Every CVSS version scores two things and combines them into a Base score: how easy the vulnerability is to reach and trigger (its exploitability), and how much damage it causes once triggered (its impact). Each metric has a small set of possible values, each carrying a numeric weight defined in the specification; the version's formula combines those weights into the 0 to 10 Base score and a matching qualitative severity rating. The Base score alone assumes a worst-case, generic deployment, it does not know whether the vulnerability is actively exploited or how critical your specific system is, that is what the other two metric groups are for.
The three metric groups
Base
The intrinsic, unchanging characteristics of a vulnerability: how it is reached and exploited, and what it impacts. Required in every version; this is the number most often quoted as "the CVSS score."
Temporal / Threat
Characteristics that change over the vulnerability's lifetime, mainly whether working exploit code exists in the wild. Called Temporal in v2.0 through v3.1 (Exploit Code Maturity, Remediation Level, Report Confidence) and narrowed to a single Threat metric, Exploit Maturity, in v4.0.
Environmental
Optional, organization-specific context: how important Confidentiality, Integrity, and Availability actually are for your deployment, plus the ability to override any Base metric to reflect compensating controls. Lets two organizations score the same vulnerability differently based on their own environment.
Version by version
4 versions are in active use. Newest first:
CVSS v4.0
2023Metric groups: Base is split into Exploitability (Attack Vector, Attack Complexity, the new Attack Requirements, Privileges Required, User Interaction) and two separate Impact groups, Vulnerable System (VC/VI/VA) and Subsequent System (SC/SI/SA), replacing Scope. Threat (Exploit Maturity only, replacing the old Temporal group's 3 metrics). Environmental (Security Requirements plus modified versions of every Base metric). Supplemental metrics (Safety, Automatable, Recovery, Value Density, Vulnerability Response Effort, Provider Urgency) are new, informational, and never change the numeric score.
The biggest change since v2.0 to v3.0. Scope is gone, replaced by scoring impact to the vulnerable system and any subsequent system separately. A new Attack Requirements (AT) metric captures preconditions beyond privileges and interaction. User Interaction gains Passive and Active values instead of just None/Required. New naming makes it explicit which metric groups were actually scored: CVSS-B (Base only), CVSS-BT (Base + Threat), CVSS-BTE (Base + Threat + Environmental) - so a bare score can no longer be mistaken for one that accounted for real-world exploit activity or organizational context.
Read the CVSS v4.0 specificationCVSS v3.1
2019Metric groups: Identical metric groups and metrics to v3.0: no new or removed metrics.
A clarification release, not a scoring redesign. It defines an explicit, integer-safe Roundup algorithm (Appendix A) to eliminate floating-point edge cases in the v3.0 rounding rule, and rewrites ambiguous metric definitions (particularly Scope and Privileges Required) for consistency across vendors. A given vector scores the same in v3.0 and v3.1 except in the rare cases where v3.0's rounding was ambiguous.
Read the CVSS v3.1 specificationCVSS v3.0
2015Metric groups: Base (Attack Vector, Attack Complexity, Privileges Required, User Interaction, Scope, and Confidentiality/Integrity/Availability impact), Temporal (Exploit Code Maturity, Remediation Level, Report Confidence), and Environmental (Security Requirements plus modified Base metrics).
A major rework of v2.0: replaced Authentication with Privileges Required, added User Interaction and the Scope metric (whether the impact extends beyond the vulnerable component), and moved to the 5-band severity scale (None through Critical) that NVD and most vendors still use today.
Read the CVSS v3.0 specificationCVSS v2.0
2007Metric groups: Base (Access Vector, Access Complexity, Authentication, and Confidentiality/Integrity/Availability impact), Temporal (Exploitability, Remediation Level, Report Confidence), and Environmental (Collateral Damage Potential, Target Distribution, and security-requirement weights).
The original FIRST specification. Scored on a 3-band qualitative scale (Low, Medium, High) rather than the 5-band scale later versions use, had no concept of Scope, and its Authentication metric counted how many times an attacker had to authenticate rather than what privilege level was required. Largely superseded, but still seen in older CVE records and some legacy vulnerability tools.
Read the CVSS v2.0 specificationSeverity bands
v3.0, v3.1, and v4.0 all map the 0 to 10 Base score to the same 5-band qualitative severity rating, the scale NVD and most vendors publish alongside every CVE. v2.0 predates Critical and None and used a narrower 3-band scale (Low, Medium, High).
| Score range | Severity |
|---|---|
| 0.0 | None |
| 0.1 - 3.9 | Low |
| 4.0 - 6.9 | Medium |
| 7.0 - 8.9 | High |
| 9.0 - 10.0 | Critical |
A brief history
CVSS originated in 2005 under the U.S. National Infrastructure Advisory Council and was handed to FIRST, which has maintained every version since. v2.0 shipped in 2007 and became the basis for NVD scoring for nearly a decade. v3.0 arrived in 2015 with a major metric rework, and v3.1 followed in 2019 as a clarification release with no new metrics. v4.0, published in November 2023, is the current version and the first to split impact into Vulnerable System and Subsequent System scoring instead of the old Scope flag. Older CVEs generally keep the score they were originally assigned rather than being retroactively rescored under the newest version.
Frequently asked questions
What is CVSS?
CVSS (Common Vulnerability Scoring System) is an open, vendor-neutral standard maintained by FIRST (Forum of Incident Response and Security Teams) for rating the severity of a software vulnerability on a 0 to 10 scale. It is the scoring system behind the severity numbers published in the NVD (National Vulnerability Database) and most CVE records.
What CVSS version should I use?
Use whichever version the vulnerability was originally scored under when comparing to published data, since scores are not always directly convertible. For new internal scoring, v3.1 is the most widely supported today (broadest tooling and vendor adoption), while v4.0 is the current standard and scores impact more precisely by separating the vulnerable system from any subsequent system it can reach.
Why did CVSS v4.0 remove Scope?
Scope in v3.x was a single yes/no flag for whether impact extended beyond the vulnerable component, which FIRST found was inconsistently applied across scorers. v4.0 replaces it with two explicit impact groups, Vulnerable System (VC/VI/VA) and Subsequent System (SC/SI/SA), so the impact to the thing that was actually exploited and the impact to anything downstream are scored separately rather than folded into one ambiguous flag.
Is a CVSS score the same as risk?
No. The CVSS Base score reflects the intrinsic, worst-case severity of a vulnerability in isolation. It does not account for whether the vulnerability is actively being exploited, whether compensating controls exist, or how critical the affected asset is to your organization, that is what the Temporal/Threat and Environmental metric groups, and ultimately your own risk process, are for.
Does TurboPentest score findings with CVSS?
TurboPentest reports findings with severity ratings, and non-source-code findings can be scored against CVSS where a CVE match applies. Source-code findings from white-box testing are classified against OWASP ASVS levels (MASVS for mobile) rather than CVSS, since ASVS is built for application-layer weaknesses that CVSS's network-oriented metrics don't fit well. Use the free CVSS calculator below to score any vulnerability by hand.
Written and reviewed by
Michel Chamberland - Founder & CEO, IntegSec
CISSP, OSCP, OSCE, CEH, GIAC, CCSK · 20+ years in offensive security
Michel has spent 20+ years on offensive security teams including IBM X-Force Red and Trustwave SpiderLabs, leading penetration tests, red team engagements, and breach response for Fortune 500 customers. He is the founder of IntegSec and the architect of TurboPentest.
Findings scored, not just listed
TurboPentest's agentic AI pentest reports severity for every finding and orders them critical to info, from $99 per target.
Start a pentest