The OWASP Risk Rating Methodology is the Open Worldwide Application Security Project’s method for rating the risk of a specific vulnerability — typically a finding from a penetration test or code review. It starts from the standard model, Risk = Likelihood × Impact, but never multiplies anything. Instead each side is estimated from eight factors, the factors are averaged, each average is banded Low, Medium or High, and a 3×3 table turns the two bands into a severity. It is a semi-quantitative method with a qualitative finish.
1 · Six Steps
- Identify the risk — the threat agent, the attack, the vulnerability and the business impact. OWASP: when there are several possible attackers, “err on the side of caution by using the worst-case option.”
- Estimate likelihood from eight factors: four about the threat agent, four about the vulnerability.
- Estimate impact from four technical factors and four business factors.
- Determine severity — average, band, and read the severity matrix.
- Decide what to fix — most severe first, and not every risk is worth fixing: if the fix costs more than the loss, accepting it can be the right call.
- Customise the model — add factors, change options, or weight factors that matter more in your business.
2 · Likelihood: Who Attacks, and How Easily
Likelihood is the average of eight ratings. The threat agent factors describe the most likely attacker group; the vulnerability factors describe the flaw itself.
| Group | Factor | Options and ratings |
|---|---|---|
| Threat agent | Skill level | No technical skills 1; Some technical skills 3; Advanced computer user 5; Network and programming skills 6; Security penetration skills 9 |
| Threat agent | Motive | Low or no reward 1; Possible reward 4; High reward 9 |
| Threat agent | Opportunity | Full access or expensive resources required 0; Special access or resources required 4; Some access or resources required 7; No access or resources required 9 |
| Threat agent | Size | Developers 2; System administrators 2; Intranet users 4; Partners 5; Authenticated users 6; Anonymous Internet users 9 |
| Vulnerability | Ease of discovery | Practically impossible 1; Difficult 3; Easy 7; Automated tools available 9 |
| Vulnerability | Ease of exploit | Theoretical 1; Difficult 3; Easy 5; Automated tools available 9 |
| Vulnerability | Awareness | Unknown 1; Hidden 4; Obvious 6; Public knowledge 9 |
| Vulnerability | Intrusion detection | Active detection in application 1; Logged and reviewed 3; Logged without review 8; Not logged 9 |
OWASP says likelihood need not be over-precise: Low, Medium or High is usually enough.
3 · Impact: Technical and Business
Impact is estimated twice. Technical impact covers the classic security properties plus accountability. Business impact covers what the business actually loses — and OWASP is explicit that it is “ultimately” the more important of the two: if good business information exists, use it instead of the technical estimate.
| Group | Factor | Options and ratings |
|---|---|---|
| Technical impact | Loss of confidentiality | Minimal non-sensitive data disclosed 2; Minimal critical data disclosed 6; Extensive non-sensitive data disclosed 6; Extensive critical data disclosed 7; All data disclosed 9 |
| Technical impact | Loss of integrity | Minimal slightly corrupt data 1; Minimal seriously corrupt data 3; Extensive slightly corrupt data 5; Extensive seriously corrupt data 7; All data totally corrupt 9 |
| Technical impact | Loss of availability | Minimal secondary services interrupted 1; Minimal primary services interrupted 5; Extensive secondary services interrupted 5; Extensive primary services interrupted 7; All services completely lost 9 |
| Technical impact | Loss of accountability | Fully traceable 1; Possibly traceable 7; Completely anonymous 9 |
| Business impact | Financial damage | Less than the cost to fix the vulnerability 1; Minor effect on annual profit 3; Significant effect on annual profit 7; Bankruptcy 9 |
| Business impact | Reputation damage | Minimal damage 1; Loss of major accounts 4; Loss of goodwill 5; Brand damage 9 |
| Business impact | Non-compliance | Minor violation 2; Clear violation 5; High profile violation 7 |
| Business impact | Privacy violation | One individual 3; Hundreds of people 5; Thousands of people 7; Millions of people 9 |
4 · From Averages to Severity
Each side is the plain average of its factors, then banded:
The two bands meet in OWASP’s severity matrix:
| Impact ↓ / Likelihood → | LOW | MEDIUM | HIGH |
|---|---|---|---|
| HIGH | Medium | High | Critical |
| MEDIUM | Low | Medium | High |
| LOW | Note | Low | Medium |
Unlike NIST’s matrix, OWASP’s is symmetric: High likelihood with Low impact and Low likelihood with High impact both come out Medium.
OWASP’s own example: threat agent ratings 5, 2, 7, 1 and vulnerability ratings 3, 6, 9, 2 average to 4.375 — Medium likelihood. Technical impact 9, 7, 5, 8 averages 7.25 — High, but business impact 1, 2, 1, 5 averages only 2.25 — Low. Technically that is a High; because the business would barely notice, OWASP concludes the overall severity is best described as Low.
5 · Worked Example: a Password-Only VPN
The series’ standard risk, framed as a finding: the remote-access VPN accepts password-only logins, and ransomware crews buy stolen VPN credentials.
| Side | Ratings | Average | Band |
|---|---|---|---|
| Threat agent | skill 6, motive 9, opportunity 7, size 9 | 7.75 | HIGH likelihood 7.75 |
| Vulnerability | discovery 9, exploit 5, awareness 9, detection 8 | 7.75 | |
| Technical impact | confidentiality 7, integrity 7, availability 7, accountability 7 | 7.00 | HIGH |
| Business impact | financial 7, reputation 5, compliance 5, privacy 7 | 6.00 | HIGH |
| Severity | HIGH likelihood × HIGH business impact | Critical |
▸ Now enforce MFA. Opportunity drops to 4 (special access needed) and ease of exploit to 3 — and likelihood only falls from 7.75 to 7.125, still HIGH. Averaging eight factors dilutes the one control that actually decides the attack. Where a single factor is decisive, override the average and say why.
Try it: OWASP risk rating
Threat agent
Vulnerability
Technical impact
Business impact
6 · OWASP’s Own Caveats
The methodology page carries an unusually frank disclaimer. It notes there has been “lots of debate” about the methodology and the weighting of threat-actor skill levels, and points readers to more established alternatives: NIST SP 800-30, the Government of Canada Harmonized TRA methodology, and Mozilla’s Risk Assessment Summary and Rapid Risk Assessment, alongside OWASP’s threat-modelling projects. Add to that:
- All factors weigh the same by default — the model “assumes that all the factors are equally important”. Step 6 allows a weighted average; most teams never use it.
- Averaging hides the decisive factor — as the MFA example shows.
- The numbers are uneven — Skill level jumps 1, 3, 5, 6, 9; two different options can share a rating. Treat the ratings as ordered labels, not measurements.
- It rates a vulnerability, not an enterprise risk — use it for findings and roll the results up into the risk register, not instead of it.
- “Quite a lot of uncertainty” — OWASP’s phrase for its own estimates. Record the reasoning behind each pick so a second tester can challenge it.
For a severity-only score that the whole industry shares, see CVSS, EPSS, KEV & SSVC; OWASP Risk Rating adds the business context CVSS leaves out.
Bottom line
- 16 factors, 0–9 each — eight for likelihood, eight for impact (four technical, four business).
- Average, band, look up — under 3 Low, under 6 Medium, else High; a symmetric 3×3 matrix gives Note to Critical.
- Business impact wins when you have it.
- Override the average when one factor decides the attack, and write down why.
- A finding-level tool — feed its results into the enterprise register.
Sources & method. Steps, factors, option ratings, bands, severity matrix, OWASP’s worked example and the quoted caveats from the OWASP Risk Rating Methodology (OWASP Foundation, community page; content licensed CC BY-SA 4.0), read in November 2026. Factor ratings are reproduced as published; the VPN example and the MFA variation are illustrative. Original analysis; no third-party graphics reproduced.
← Back to Cyber security