Skip to main content

    30 June 2026 · 7 min read

    External vs Internal Penetration Testing: Differences, Scope, and When You Need Each

    External and internal penetration tests sound similar but model two different attackers, the outsider trying to get in, and the intruder already inside. Here is how they differ and when each one matters.

    Request a quote

    What is an external penetration test?

    An external penetration test assesses your internet-facing perimeter the way a remote attacker would: from outside your network, with no prior access. The tester targets what is reachable from the public internet, your websites and web applications, VPN and remote-access gateways, mail and DNS infrastructure, exposed APIs, and any forgotten or shadow assets that should not be there.

    External vs internal penetration testing

    Two starting points, one attack path

    Internet

    • Remote attacker, no prior access

    Perimeter · external test

    • Websites and web applications
    • VPN and remote access
    • Mail, DNS and exposed APIs

    Inside · internal test

    • Assumed-breach foothold
    • Active Directory and lateral movement

    Crown jewels

    • Critical systems and data
    The external test asks whether an attacker gets in. The internal test asks how far they get once inside.

    The goal is to answer one question: can an attacker on the internet break through your perimeter and gain a foothold inside? External testing surfaces exposed services, weak authentication, unpatched edge systems, and misconfigurations that turn a public-facing asset into an entry point. For most organisations it is the natural starting point, because it mirrors how the majority of real intrusions begin.

    What is an internal penetration test?

    An internal penetration test assesses what an attacker could do once they are already inside your network, whether through a phished employee, a compromised laptop, a malicious insider, or a breached supplier. The tester starts from a position on the internal network and tries to escalate: move laterally between systems, harvest credentials, reach domain administrator, and get to your most sensitive data.

    Internal testing answers a different question: if the perimeter is bypassed, how far can the intrusion spread? It exposes flat network segmentation, weak internal credentials, over-privileged accounts, unpatched internal services, and Active Directory weaknesses, the conditions that turn a single compromised device into a full breach. It is essential for organisations that hold sensitive data or must demonstrate defence in depth.

    External vs Internal Penetration Testing, Key Differences

    Both improve your security, but they model different attackers and answer different questions. The external test asks "can someone break in?"; the internal test asks "how far can they get once inside?"

    DimensionExternal Penetration TestInternal Penetration Test
    Attacker modelledRemote outsider, no accessIntruder or insider already on the network
    Starting pointThe public internetA foothold inside your network
    Primary targetInternet-facing perimeter and assetsInternal systems, accounts, and data
    Question answeredCan an attacker break in?How far can a breach spread once inside?
    Typical findingsExposed services, weak edge auth, unpatched gatewaysLateral movement, privilege escalation, weak segmentation
    Best first forMost organisations (mirrors real intrusions)Sensitive-data holders and defence-in-depth needs

    When should you run which?

    For most organisations the answer is not either-or. The two are complementary halves of a realistic assessment. Use this as a guide.

    Run an external penetration test when

    You want to validate your internet-facing perimeter, are launching or changing public services, need to know what an outside attacker can reach, or are starting a security programme and want to begin where most real intrusions begin.

    Run an internal penetration test when

    You hold sensitive data, must demonstrate defence in depth for compliance, want to understand blast radius after a phishing or supplier compromise, or have a mature perimeter and need assurance that one foothold does not become a full breach.

    Run both together when

    You want a realistic picture of the whole attack path, from initial access through to impact. External testing proves whether the perimeter holds; internal testing proves what happens if it does not. Together they map the route a real adversary would take end to end.

    Common findings in an external penetration test

    These are issues testers regularly find on internet-facing systems.

    • Exposed admin interfaces, test environments, or forgotten subdomains
    • Unpatched VPN gateways, firewalls, and remote access services
    • Weak or reused passwords on login portals without multi-factor authentication
    • Misconfigured cloud storage or services reachable from the internet
    • Web application flaws such as broken access control and injection
    • Error messages, headers, or public files that leak information an attacker can use to plan

    Common findings in an internal penetration test

    Once inside the network, testers look for paths that turn one foothold into wider access.

    • Active Directory misconfigurations that allow privilege escalation
    • The same local administrator password reused across many machines
    • Flat networks with no segmentation between user devices, servers, and sensitive systems
    • Legacy network protocols that expose credentials
    • File shares holding passwords, keys, or sensitive data
    • Unpatched internal servers that are hidden from the internet but reachable from any employee device

    How to scope each test

    A clear scope keeps the test focused and makes the estimate accurate. This is what testers typically need from you.

    External test scope

    The domains, IP ranges, and cloud services you own, any third-party hosted applications (with the host's permission where required), and which login portals and APIs are in scope.

    Internal test scope

    The starting point (an office network connection, a VPN account, or a standard employee laptop), the sites and network segments in scope, and any systems that must be excluded or tested only at agreed times.

    Rules of engagement for both

    Testing windows, emergency contacts, how critical findings are reported during the test, and whether your monitoring team is told in advance or the test is also used to check detection.

    Frequently asked questions

    Not sure which scope you need?

    We will help you decide and scope the right test, external, internal, or both, based on your environment and risk. Talk to our penetration testing team.