Penetration testing should be conducted regularly across all critical systems, with additional testing after significant changes. Systems that handle sensitive data, are exposed to the internet, or undergo frequent updates require more frequent assessments or continuous testing. Compliance standards provide a baseline, but the frequency of changes within your environment should ultimately determine the testing schedule.

Most companies schedule penetration tests around compliance audits. When an audit approaches, testing is conducted, and the report is filed as evidence. But attackers do not follow audit schedules. A redesigned login flow, a cloud migration, or a major infrastructure update can introduce new vulnerabilities that remain undetected until the next scheduled test. This gap is the real concern behind penetration testing frequency.

How Often Should Penetration Testing Be Done?

Organisations should conduct penetration testing at least once a year for every infrastructure in scope and repeat the testing whenever they implement a significant change. That is the baseline that most standards and most security teams agree on. The right cadence for your company then depends on risk and on how fast your systems change.

SituationRecommended penetration testing frequency
Stable internal system, low exposureEvery 12 months, and after significant changes
Internet-facing application or APIEvery 12 months at minimum, plus a test after each major release
Sensitive data: payments, health, financialEvery 6 months, and after significant changes
Weekly or daily releasesContinuous testing, with a deep manual test at least once a year
Regulated entity with a set cadenceWhatever the regulator sets, as a floor
After a breach or serious incidentImmediately, once the incident is contained

Treat the table as a starting point. Two companies in the same industry can need very different schedules. One may ship code daily while the other changes its systems twice a year.

Why Annual Penetration Testing Is a Floor, Not a Target

Annual penetration testing means a full penetration test of each system in scope at least once every 12 months.  NIST’s technical guide to security testing, NIST SP 800-115, is still NIST’s reference on the subject, 18 years after publication. The guide notes that, given the cost and risk of penetration testing, annual testing may be enough for many organisations. Most standards that set a number pick the same 12 months.

Think about how a car gets serviced. One rule says once a year. Another says every 10,000 kilometres. A car that sits in the garage most of the year is fine with the calendar rule. A car that drives 50,000 kilometres a year will break down long before its annual service.

Your applications work the same way. The calendar tells you when a test is due. The change log, your odometer, tells you when one is overdue.

That is why annual testing is a floor. It suits stable systems with low exposure. For anything that changes often, an annual test describes a system that no longer exists by the time the next one starts. Read whether your annual pentest report is enough for an industry-by-industry view of that gap.

Run a Pentest After Major Changes

A penetration test after major changes is the part of the schedule most companies skip. PCI DSS v4.0.1 requires one after “any significant infrastructure or application upgrade or change”. The same logic applies to every system that matters, regulated or not.

Changes that should trigger a new penetration test include:

  • New applications, APIs, or products going live, or an existing system opening to the internet for the first time
  • Changes to authentication or authorisation: new login flows, single sign-on, multi-factor authentication, or role models
  • Changes to payments or other sensitive data flows
  • Cloud migrations, or major changes to cloud architecture or identity and access management
  • New network segmentation, or changes to existing segmentation controls
  • Mergers and acquisitions that connect new systems to yours
  • Major third-party integrations with access to your data
  • Significant breaches or incidents, once they are contained

A useful rule of thumb: if the change would need a fresh architecture review, it needs a fresh penetration test. Routine patches and small features usually do not. A rebuilt login page usually does.

Scope the change-triggered test to the change and to what it touches. A new payment API needs testing of the API, the services it calls, and the access controls around it, not the entire estate. That keeps event-driven testing affordable, so it actually happens.

Internal vs External Penetration Testing Frequency

Internal and external penetration tests answer different questions, and they do not always need the same cadence. An external test looks at what an attacker on the internet can reach: websites, APIs, VPN gateways, email systems, and cloud services. An internal test assumes the attacker is already inside, perhaps through a phished laptop or a compromised supplier account. The question is how far they can move.

External systems face constant automated probing, so most companies test them at least once a year and after every significant internet-facing change. Many test their most exposed applications every 6 months. Many companies change their internal networks more slowly. For them, an annual internal test is a common baseline, plus a test after major network or identity changes.

PCI DSS does not let you choose one over the other. It requires both internal and external penetration testing at least every 12 months and after significant changes. If you only budget for one, the external test usually comes first. But only an internal test shows whether one stolen password can become a full breach.

Penetration Testing Frequency by Compliance Standard

Compliance rules for pentesting vary more than most guides admit. Some standards set a clear number. Others leave the cadence to your risk assessment, and several popular articles claim rules those standards do not contain.

StandardSets a pentest frequency?Requirement
PCI DSS v4.0.1YesInternal and external tests at least every 12 months and after significant changes; segmentation tests every 12 months, or every 6 months for service providers
RBI IT Master Direction (2023)YesPenetration testing at least every 12 months for critical and DMZ systems; vulnerability assessment every 6 months
SEBI CSCRF (2024)YesVAPT yearly for most regulated entities; half-yearly for Qualified Stock Brokers and NCIIPC-designated protected systems
CERT-In audit guidelines (2025)Yes, for auditsAt least one cyber security audit a year; solely tool-based testing discouraged
HIPAA Security RuleNot yetNo pentest rule today; a 2025 proposal would require one every 12 months
ISO/IEC 27001:2022NoTesting cadence set by your own risk assessment
SOC 2NoPenetration testing is one way to evaluate controls, not a fixed rule
NIST SP 800-53 (Rev. 5)NoControl CA-8 requires penetration testing at a frequency your organisation defines; part of the high-impact baseline

PCI DSS

PCI DSS is the most specific. Requirement 11.4 calls for internal and external penetration testing at least once every 12 months and after any significant infrastructure or application change. It also requires that exploitable findings are corrected and the test repeated to verify the corrections.

Segmentation controls that isolate the cardholder data environment need their own tests. Merchants test them at least every 12 months, and service providers every 6 months. This summary of PCI DSS requirement 11 sets out the full text, and this PCI DSS compliance guide covers the wider standard.

RBI, SEBI, and CERT-In

Indian regulators set clear numbers too. Clause 26 of the RBI’s IT Master Direction covers critical systems and customer-facing systems in the DMZ. Both need penetration testing at least every 12 months and vulnerability assessment every 6 months. Banks and NBFCs can map the wider rules with this RBI audit checklist.

SEBI’s Cybersecurity and Cyber Resilience Framework splits by category. Exchange circulars such as this NSE VAPT circular set yearly for most regulated entities. Qualified Stock Brokers, and systems NCIIPC has designated as protected or critical infrastructure, test half-yearly. Both must use a CERT-In empanelled auditor. For the timelines and closure rules, see this breakdown of SEBI’s VAPT guidelines.

CERT-In’s 2025 audit guidelines set a baseline of at least one cyber security audit a year, which sectoral regulators can raise. The same guidelines discourage testing that is solely tool-based, so an automated scan will not satisfy the rule on its own. Insurers follow their own regulator’s rules, summarised in this overview of IRDAI’s 2026 cybersecurity guidelines.

ISO 27001, SOC 2, NIST 800-53, and HIPAA

ISO/IEC 27001:2022 does not set a penetration testing frequency. Its controls on technical vulnerability management and security testing require you to decide the cadence through your own risk assessment, then follow it. Most certified companies test at least annually because auditors expect evidence that the controls work. The wider rules are covered in this guide to ISO 27001 compliance.

SOC 2 works the same way. The Trust Services Criteria treat penetration testing as one way to evaluate whether controls operate effectively, not as a fixed rule. In practice, most SOC 2 auditors expect at least one test a year, as this guide to SOC 2 compliance explains.

NIST SP 800-53 takes the same approach. Its NIST control CA-8 requires penetration testing “at an organisation-defined frequency” on systems you choose. The control sits in the high-impact baseline, not the moderate one. Organisations that follow it usually set the cadence through their own risk assessment, often at once a year.

HIPAA is the one most often misreported. The current Security Rule requires periodic evaluation of security controls but does not mention penetration testing.

In January 2025, HHS proposed a rule that would require penetration testing at least once every 12 months. It has not been finalised, and law firm McDermott Will & Emery reports that final action is projected for July 2027. Healthcare companies should plan for annual testing anyway, since the proposal signals where the rule is heading. This HIPAA compliance checklist covers the current rules.

Factors That Decide Your Penetration Testing Frequency

Compliance gives you a floor. The factors below decide how far above that floor your schedule should sit.

Rate of change. The more often a system changes, the faster a test goes out of date. A system released weekly needs more than a yearly test, while a system that changes twice a year may not.

Exposure. Internet-facing applications, APIs, and remote access systems face constant probing. Internal systems behind several layers of control face less.

Data sensitivity. Systems that hold payment data, health records, financial records, or personal data at scale carry a higher cost of failure. They justify more testing.

Past findings. If the last test found critical issues, or fixes took months to close, test again sooner. A clean result with fast remediation supports a longer interval.

Threat activity. A new class of attack against your technology stack, or an incident at a peer company, is a reason to test early.

Business events. Mergers, new markets, and major customer contracts often bring new systems, new integrations, and new testing obligations.

You do not need the same cadence for every system. Most companies do better by sorting systems into tiers:

TierTypical systemsTesting cadence
Tier 1Payment, login, customer data, internet-facing core applicationsEvery 6 months, after every significant change, with continuous testing between
Tier 2Internal business applications, partner portalsEvery 12 months and after significant changes
Tier 3Low-risk internal tools with little sensitive dataEvery 12 to 24 months, or as standards require

Penetration Testing Frequency by Industry

Industry shapes penetration testing frequency mainly through the regulators and the data involved. The schedules below follow from each sector’s regulatory rules, plus the exposure typical of that sector.

IndustryCommon scheduleWhat drives it
Banking and NBFCsEvery 12 months for critical systems, vulnerability assessment every 6 months, plus tests after changesRBI’s IT Master Direction
Capital marketsYearly VAPT, or half-yearly for Qualified Stock Brokers and protected systemsSEBI’s cyber resilience framework
Payments and e-commerceEvery 12 months and after significant changes; segmentation every 6 months for service providersPCI DSS
HealthcareAt least every 12 months is the safe planHIPAA’s proposed rule and the sensitivity of health data
SaaS and technologyEvery 12 months for SOC 2 and customer contracts, with continuous testing for weekly releasesCustomer due diligence and release speed
Manufacturing and operational technologyEvery 12 months, scoped and timed carefullyProduction systems where testing itself carries operational risk

The table shows the common pattern, not a rule. A small fintech that releases daily needs more frequent testing than a large bank’s rarely changed back-office system, whatever the sector averages say.

Building a Compliance Testing Schedule That Holds

A compliance testing schedule for penetration testing works best when it is built around systems and changes, not around audit dates. Start by listing every system in scope, its tier, and every standard that applies to it. Then take the strictest cadence for each system, so one schedule satisfies PCI DSS, RBI, SEBI, and your own risk assessment at once.

Next, plan the year around 3 kinds of test:

  • Scheduled tests at the cadence each tier needs, spread across the year so that results and remediation do not all land in the same month
  • Change-triggered tests for significant changes, with a standing agreement that lets one start within days
  • Retests after each fix, because standards such as PCI DSS require repeated testing to verify corrections, and SEBI expects closure within 3 months with revalidation

Here is how that looks for an illustrative mid-size payments company with a customer app, a merchant API, and an internal back office:

QuarterPlanned penetration testing
Q1External test of the customer app and merchant API; segmentation test of the cardholder data environment
Q2Retests of Q1 findings; change-triggered test for a new login flow
Q3Internal network test; second segmentation test, required because the company is a service provider
Q4Second test of the customer app and merchant API, since both are Tier 1; retests before year end

The pattern matters more than the exact quarters. Spread the tests across the year, and let the strictest standard drive each system. Leave room for changes nobody planned in January.

Set aside part of the budget for change-triggered tests. Organisations are most likely to cut these tests from their plans, even though they are also most likely to uncover serious flaws. The article Remediation Over Reports explains why organisations should invest in retesting.

Signs Your Penetration Testing Frequency Is Too Low

The clearest evidence that you test too rarely is in your own results. Watch for these signs:

  • The same findings come back. If last year’s critical issues reappear, fixes are not holding, and a longer gap only lets them settle in.
  • Every test finds something critical. A steady stream of critical findings means new risk is arriving faster than your testing cycle can catch it.
  • Major releases ship untested. If you can name a significant change from the past 6 months that no tester has looked at, your schedule is behind your systems.
  • Testing follows the audit, not the change log. A schedule set by audit dates alone will always leave the riskiest months uncovered.
  • Retests never happen. A finding that is marked fixed but never retested is still a finding.

Any 2 of these together usually mean moving your highest-risk systems to a 6-month cycle, or adding continuous testing between manual engagements.

How Often Should Vulnerability Scans Be Performed?

Vulnerability scans should run far more often than penetration tests. A scan is an automated check for known weaknesses, so it is fast and cheap to repeat. A penetration test is skilled manual work, so it is slower and costs more. The two cadences should be planned together, not confused.

Standards set the minimum for scanning too. PCI DSS requires internal and external vulnerability scans at least once every 3 months. External scans must come from an Approved Scanning Vendor, with rescans until the results pass. The RBI expects vulnerability assessment of critical systems at least every 6 months.

Good practice goes further than those minimums:

  • Internet-facing systems: weekly or continuous scanning, plus a scan after every deployment
  • Internal systems: at least monthly, plus a scan after significant changes
  • New critical vulnerabilities: a targeted scan within days of public disclosure

A clean scan does not replace a penetration test. Scanners find known weaknesses, while testers find what scanners cannot, such as broken access control and chained attacks. Plan both cadences in the same schedule.

Where Continuous Testing Fits

Continuous testing changes the frequency question. Automated testing runs on a schedule and after each deployment, while manual testers return to critical systems at set intervals and after significant changes. The result is that no change waits a year to be tested. The full case is made in this post on why continuous pentesting matters more than annual assessments.

Continuous testing does not replace the deep manual test. Automated tools cannot find most access control flaws, business logic issues, or chained attacks, as this comparison of automated and manual pentesting shows. The two work best together: automation for constant coverage, manual testing at the cadence your tiers need.

A penetration testing as a service (PTaaS) model packages both into one ongoing engagement. A change-triggered test then becomes a request on a platform, not a new procurement cycle. Read what PTaaS is for how the model works. Once your testing programme is mature, add red teaming exercises to test whether your teams detect and stop a realistic attack.

Kratikal’s PTaaS supports this kind of schedule. The service offers on-demand manual tests after significant changes, retests on request, and continuous automated coverage through AutoSecT.

Setting Your Penetration Testing Frequency

The audit calendar is easy to follow. The change log is harder, but it is the schedule attackers follow. Set penetration testing frequency at 12 months as the floor for every system. Move your highest-risk systems to 6 months, and test everything that matters after every significant change.

Then write it down as a schedule by system and tier, using the strictest rule that applies to each. Budget for change-triggered tests and retests, not just the annual ones.

If you are planning that schedule now, Kratikal’s PTaaS testing services can run tests on demand after each change. Its VAPT testing services cover the scheduled engagements regulators expect.

FAQs

How Often Should Penetration Testing Be Done?

Penetration testing should be done at least once every 12 months for every system in scope. Test again after any significant change, such as a new application, a rebuilt login flow, or a cloud migration. High-risk systems that handle sensitive data or face the internet often need testing every 6 months. Systems released weekly benefit from continuous testing, with a deep manual test at least once a year.

What Is the PCI DSS Penetration Testing Frequency?

The PCI DSS penetration testing frequency is at least once every 12 months for both internal and external testing. A new test is also due after any significant infrastructure or application upgrade or change. Requirement 11.4 of PCI DSS v4.0.1 also requires that exploitable findings are corrected and the testing repeated to verify the corrections. Segmentation controls have their own testing schedule.

How Often Should Organisations Perform Penetration Tests for Segmentation Controls Under PCI DSS?

Under PCI DSS, organisations must conduct penetration tests for segmentation controls at least once every 12 months and after any change to segmentation controls or methods. Service providers must follow a stricter schedule and conduct these tests at least once every 6 months. Organisations must test all segmentation controls in use and confirm that they effectively isolate the cardholder data environment from out-of-scope systems.

What Is the ISO 27001 Penetration Testing Frequency?

ISO/IEC 27001:2022 does not set a fixed penetration testing frequency. Its controls on technical vulnerability management and security testing require you to set a cadence through your own risk assessment and then follow it. In practice, most certified companies conduct penetration tests at least once a year because auditors look for evidence that organisations regularly test their security controls and verify their effectiveness.

When Is Penetration Testing Required?

Organisations must conduct penetration testing when applicable standards or regulators require it. Examples include PCI DSS, the RBI’s IT Master Direction, and SEBI’s cyber resilience framework. Organisations should also conduct penetration tests after significant changes to systems in scope, following a security incident, and when customer contracts require them. Standards such as ISO 27001 and SOC 2 allow organisations to determine the testing frequency based on their risk assessments.

Does NIST 800-53 Require Penetration Testing?

NIST SP 800-53 includes penetration testing as control CA-8, but it does not set a frequency. The control asks you to conduct penetration testing at an organisation-defined frequency on systems you choose. CA-8 is part of the high-impact baseline, not the moderate one. Many organisations that follow NIST 800-53 still test at least once a year, based on their own risk assessment.