Burp Suite is the most widely used intercepting proxy for web application testing, and in the wrong configuration it will happily send traffic to systems nobody authorised you to touch. This guide is for defenders and authorised testing only. It explains what Burp Suite is, how its editions and licensing differ, how to configure scope, projects, extensions and out-of-band testing so that an engagement stays inside its written authorisation, and what a security team sees when Burp is pointed at its applications. It is written for application security engineers, internal testers, test managers and the SOC and platform teams who govern tool use. It does not teach attack techniques: for testing method, see the Web Application Security Testing Guide.
This is an independent guide from NHI Mgmt Group, based on PortSwigger's public documentation as of November 2026. It is not sponsored or reviewed by PortSwigger. Product features change, so confirm settings against PortSwigger's current documentation.
Key takeaways
- Authorisation comes before installation. PortSwigger's own documentation warns that testing can damage target systems and that you should not test systems the owner has not authorised, so written scope and rules of engagement are the first control.
- Scope is a technical control, not just a document. Target scope in Burp can restrict proxy logging, interception and live scanning to in-scope hosts; configure it from the signed scope before the first request.
- Project files are sensitive data. A disk-based project holds every request and response captured, including session tokens, credentials and personal data, so it needs encryption, access control and a retention date.
- Extensions and Collaborator change your trust boundary. Third-party extensions run inside the tool with access to all captured traffic, and out-of-band testing makes targets talk to a server outside the application, so both need explicit decisions.
- Defenders should expect to see it. Authorised Burp testing is visible in WAF, web server and DNS logs, which makes it a useful check of detection coverage when the SOC is told in advance.
What the tool does
Burp Suite, from PortSwigger, sits between a browser and a web application as an intercepting HTTP proxy. The tester's traffic flows through Burp Proxy, where requests and responses can be viewed, held and modified, and every exchange is recorded in HTTP history and the site map. Around that core sit manual tools (Repeater for re-sending individual requests, Intruder for structured request variation, Comparer, Decoder and others), an automated vulnerability scanner (Burp Scanner), and Burp Collaborator, a service for detecting vulnerabilities that only reveal themselves through out-of-band interactions. The tool can be extended through the BApp Store, custom extensions, Bambdas and BChecks.
Editions and licensing
| Edition | How PortSwigger describes it | Practical differences that matter for governance |
|---|---|---|
| Burp Suite Community Edition | Free manual tools to start web security testing | No licence key; temporary in-memory projects only (data is lost when Burp closes); no Burp Scanner; no Collaborator for manual testing |
| Burp Suite Professional | Desktop web penetration testing toolkit, per-user licence | Licence key entered at first launch; disk-based projects that save incrementally; Burp Scanner including live in-scope auditing; Collaborator for manual and automated checks |
| Burp Suite DAST | Enterprise dynamic web vulnerability scanner for automated scanning at scale | Server product deployed as Cloud or Self-hosted (standard instance, Kubernetes, or CI-driven scans with no dashboard); organisation users added centrally; Collaborator used by scan checks |
Two licensing points are easy to miss. First, a Professional licence is tied to a named user, so shared keys on a team jump box break both the licence terms and your audit trail. Second, Community Edition being free does not make it ungoverned: it captures exactly the same sensitive traffic as Professional, it simply loses it at the end of the session. Check current commercial terms with PortSwigger directly, as this guide does not reproduce pricing.
Defender use cases
Burp is not only a tool for external penetration testers. Inside an organisation it has several legitimate, defensive uses, each of which still needs authorisation from the application owner.
| Use case | Who uses it | Typical edition | Control it supports |
|---|---|---|---|
| Manual testing before release of high-risk features (authentication, payments, file upload) | AppSec engineers, internal testers | Professional | NIST SP 800-53 SA-11 Developer Testing and Evaluation |
| Scheduled or pipeline dynamic scanning of web apps and APIs | Platform and DevSecOps teams | DAST | RA-5 Vulnerability Monitoring and Scanning; CSF 2.0 ID.RA-01 |
| Independent penetration testing under contract | Third-party testers | Professional | CA-8 Penetration Testing; CIS Control 18 |
| Reproducing and validating a reported vulnerability or bug bounty submission | AppSec triage | Community or Professional | ID.RA-01 (validate and record vulnerabilities) |
| Checking that a WAF rule or code fix actually blocks the reported request | AppSec and WAF owners | Community or Professional | Fix verification before closing a finding |
| Purple team exercises measuring web attack detection | Red and blue teams together | Professional | CA-8(2) Red Team Exercises; DE.CM-01 |
| Developer training in a lab against deliberately vulnerable apps | Developers | Community | Secure coding awareness |
For how manual testing and automated scanning fit together, see the Penetration Testing Guide and the Vulnerability Scanning Guide. For pipeline use, see the DevSecOps Guide.
Deployment and safe use
Legal and authorisation requirements
- Written authorisation from the system owner: In the UK, accessing computer material without authorisation is an offence under section 1 of the Computer Misuse Act 1990, and section 3 covers unauthorised acts that impair the operation of a computer. Section 3A covers making, supplying or obtaining articles for use in those offences, which is why the purpose and authorisation of tool use should be documented. Other jurisdictions have equivalents (for example the US Computer Fraud and Abuse Act). Get legal advice for cross-border or third-party-hosted targets.
- Third-party hosting and SaaS: The application owner may not own the infrastructure. Cloud providers, CDNs and SaaS platforms publish their own testing policies; confirm the target is covered before testing. See the Penetration Test Scoping Checklist.
- Rules of engagement: Testing windows, rate limits, prohibited actions (for example no automated scanning of production payment flows), emergency contacts and a stop condition. The Red Team Rules of Engagement Template covers the structure.
Safe configuration
| Area | Setting or feature | Safe practice |
|---|---|---|
| Scope | Target > Scope, with Use advanced scope control where needed | Enter only the hosts, ports and paths in the signed scope. Advanced scope matches on protocol, host or IP range, port and file path, and every field in a rule must match. Use Include subdomains only when subdomains are explicitly in scope, as PortSwigger notes it can greatly widen scanning |
| Proxy logging | Log and intercept only in-scope items | Reduces the chance of capturing personal or third-party data from unrelated sites the browser contacts |
| Scanning | Live audit of in-scope requests (Professional) | Enable only after scope is set and only in agreed windows; prefer non-production or a dedicated test tenant |
| Network path | Settings > Network > Connections: Upstream proxy servers, SOCKS proxy, Timeouts | Route testing through an agreed egress IP so the client can identify and allow-list or monitor test traffic |
| Projects | New project on disk (Professional) | One project per engagement, stored on an encrypted volume; PortSwigger notes projects can grow to several gigabytes |
| Imported settings | Project and configuration file import | Import only from trusted sources, as PortSwigger recommends; a shared configuration can silently widen scope |
| Extensions | BApp Store and custom extensions | Maintain an approved extension list; PortSwigger reviews BApp Store submissions but makes no warranty about third-party extensions |
| Out-of-band testing | Burp Collaborator (public or private server) | Decide per engagement whether target interactions may reach the public Collaborator service; a private Collaborator server keeps that data inside your control |
Data handling
Treat every Burp project as containing live credentials and personal data. Requests carry session cookies, bearer tokens, API keys and form data, and responses may carry customer records. Practical rules: test with dedicated test accounts and synthetic data, never with a real customer account; store projects only on encrypted, access-controlled storage owned by the engagement; strip or redact tokens and personal data from evidence before it goes into a report; rotate or revoke the test accounts and any API keys used once testing ends; and agree a deletion date for raw project files in the contract. Where personal data is processed, UK GDPR Article 32 security obligations apply to the testing data too. Test credentials are non-human identities in their own right, so give them an owner, a lifetime and a revocation step. See the Secrets Management Guide.
Rollout of Burp in an organisation
- Define who may use it: named roles (AppSec, approved testers), with Professional licences issued per person and recorded in the software asset register.
- Build a standard baseline: a reviewed project configuration with logging limited to scope, an approved upstream proxy, and the approved extension list. Version it.
- Pilot in a lab: run the baseline against a deliberately vulnerable application in an isolated lab first. See the Building a Cybersecurity Home Lab Guide.
- Tie each use to a ticket: every engagement references an authorisation record that lists scope, window, test accounts and the SOC notification.
- Extend to DAST: only after manual use is governed, add scheduled or CI-driven scans against non-production environments, then production with throttled scan configurations agreed with service owners.
- Roll back: if a scan degrades service, the rules of engagement name who stops it and how; pause the DAST schedule or revoke the scan credentials, then review scope before resuming.
Detection and hardening value
When Burp is used against your applications, authorised or not, the evidence appears in familiar places. Authorised tests are an opportunity to check that it does.
| Telemetry source | What testing typically looks like | Related ATT&CK or D3FEND |
|---|---|---|
| WAF and reverse proxy logs | High volumes of requests from one source with systematically varied parameters, many 4xx and 5xx responses, rule hits across injection categories | T1595.002 Vulnerability Scanning; T1190 Exploit Public-Facing Application; D3-ITF Inbound Traffic Filtering |
| Web server and application logs | Requests to unlinked paths, unusual methods, malformed input, error stack traces | T1190; D3-NTA Network Traffic Analysis |
| DNS and egress logs from application servers | Application hosts resolving or connecting to unfamiliar external domains, a sign of out-of-band testing (or real exploitation) | D3-NTA |
| Authentication logs | Repeated failed logins or token reuse against test accounts | Account monitoring |
Metrics that show the programme is working:
- Percentage of Burp engagements with a matching authorisation record and SOC notification (target 100%).
- Percentage of authorised test activity that generated at least one SOC alert or WAF detection within the agreed time.
- Mean time to remediate findings from manual and DAST testing, by severity.
- Percentage of internet-facing applications tested in the last 12 months.
- Number of unapproved extensions or unlicensed installations found in software inventory.
Unexplained out-of-band interactions from production servers deserve investigation even if a test is running: confirm the test schedule and source before closing the alert. See the Web Application Firewall Guide and the Purple Teaming Guide.
Common failure patterns
- Scope left open: no target scope set, so proxy history and live scanning pick up every host the browser touches, including third-party analytics and identity providers.
- Testing production with real accounts: captured tokens belong to real users and the project file becomes a breach waiting to happen.
- Project files on laptops and file shares: unencrypted projects outliving the engagement by years.
- Unvetted extensions: a convenient extension from an unknown source with full access to captured traffic.
- SOC not told: authorised tests trigger an incident response, or worse, a real attack is dismissed as "probably the pen test".
- Unthrottled automated scans: scans against fragile production endpoints create records, send emails or degrade service.
Alternatives
| Option | Type | When it fits better |
|---|---|---|
| OWASP ZAP | Open source intercepting proxy and scanner | No licence budget, heavy automation or CI use, or a preference for open source. See the OWASP ZAP Guide |
| Commercial DAST platforms | Hosted or self-hosted scanners | Portfolio-wide scheduled scanning with central reporting. See the Application Security Testing Buyer's Guide |
| Browser developer tools | Built-in inspection | Quick inspection of requests without a proxy; no scanning or replay at scale |
| Third-party penetration testing provider | Service | No in-house skills, or independence required. See the Penetration Testing Provider Buyer's Guide |
Smaller organisations usually get most value from a commissioned annual test plus ZAP or DAST in the pipeline; larger organisations with in-house AppSec teams justify Professional licences for manual work and DAST for coverage. The wider tool landscape is covered in the Open Source Penetration Testing Tools Landscape Guide.
Practitioner checklist
- Signed authorisation from the system owner (and hosting provider where required) on file before any traffic is sent.
- Target scope in Burp configured from the signed scope, with logging and interception limited to in-scope items.
- Dedicated test accounts and synthetic data; no real customer accounts.
- Disk-based projects stored encrypted, with an owner and a deletion date.
- Approved extension list enforced; imported configurations only from trusted sources.
- A per-engagement decision on public versus private Collaborator.
- SOC notified with source IPs, window and contact; detections reviewed afterwards.
- Test credentials and tokens revoked when testing ends; evidence redacted before reporting.
Standards and references
- PortSwigger: Burp Suite documentation
- PortSwigger: Target scope
- PortSwigger: Burp projects
- PortSwigger: Burp Collaborator
- PortSwigger: Burp extensions
- PortSwigger: Burp Suite DAST
- OWASP: Web Security Testing Guide
- UK legislation: Computer Misuse Act 1990
Related NHI Mgmt Group resources: Web Application Security Testing Guide · OWASP ZAP Guide · Penetration Testing Guide