Practical guides · Practical guide

UAT vs PQ: when can user acceptance tests support GxP evidence?

Users have accepted the system while GxP release remains open. Identify which critical process steps have been demonstrated and under which conditions. This may reveal a gap in the evidence or its interpretation before you commission more tests.

Table of contents

At a glance

Can a user acceptance test serve as PQ?

Appropriate UAT evidence can support testing of the intended routine process where requirements, risks, version, test data, acceptance criteria and outcomes match. Business acceptance alone is insufficient. Assess what each test demonstrates and add any missing GxP cases. Renaming a UAT record does not change its evidence.

Acceptance and suitability are distinct issues.

UAT stands for User Acceptance Test. During this process, users typically verify whether a system supports the agreed business requirements. PQ is used in the context of qualification to provide evidence that the intended application functions appropriately under representative conditions. The precise terminology depends on the system, process, and quality management system.

A UAT can be well-documented and technically thorough. However, it may also focus primarily on usability, training requirements, or contractual acceptance criteria. Therefore, its value for GxP documentation cannot be inferred from the term alone. Conversely, a PQ title does not render a superficial test credible.

For computerized GMP applications, Annex 11 is the specific reference. Annex 15 assists in the assessment of qualification and representative operating conditions. This does not imply a universal requirement to establish exactly three documents titled IQ, OQ, and PQ for every software.

Sources: [1] [2]

Describe the intended routine process in full.

Describe what the future user does from the start until the decision point. A batch-specific laboratory application may include order receipt, sample assignment, result entry, review, release, and handover. Document the roles and relevant data transfers. This constitutes the reference framework for the test statement.

Afterward, assess which exceptions are professionally relevant: missing value, incorrect assignment, discrepant result, representation, locked batch, or interrupted interface. A smooth normal case alone does not demonstrate that the procedure can adequately handle critical deviations.

Define representative conditions. This may include data volumes, role distributions, device type, location, or connected systems. "Production-like" without these attributes is an ambiguous description. Complete production data is not automatically required for testing; appropriate controlled test data can effectively simulate the required scenarios.

Sources: [1] [2]

Depict existing UAT cases on their test statement

Assign each relevant UAT case to the requirement and the associated risk. Review the actual execution status, data, and results. "User satisfied" is not a precise indicator of whether a batch is blocked due to incomplete review. However, documented blocking with an appropriate expected outcome can support this assertion.

Distinguish test execution from user release. A signed acceptance protocol may consolidate multiple results. For GxP evaluation, individual information items must remain retrievable. If only a single overall score is provided, the associated supporting documentation may be missing.

Additionally, evaluate changes since the UAT. New roles, limits, or interfaces may affect previous statements. Only adopt what remains valid within the current scope. The enhancement should address the specific gap, not automatically replicate all UAT cases.

Using UAT findings to support a GxP assessment
Existing UAT contentOpen questionAppropriate complement
Task is created Is assignment to the correct batch reliable? Targeted review of incorrect and missing assignments
Result is displayed Are unit, calculation, and limit decision correct? Independent expected values, including the limit
Release button functions Can this role approve and are the prerequisites met? Positive and negative roll/condition cases
The report appears correct. Is the decision-relevant information complete? Comparison of content with reference dataset
Interface successfully sends What happens in case of failure and repetition? Failure and Retry Incidence

Sources: [1]

Example: Laboratory results until quality decision

During UAT, a result of 9.8 mg is recorded and deemed plausible, and is subsequently forwarded. The agreed upper limit is 10.0 mg inclusive. The test initially confirms that the normal workflow is executable; it does not yet address questions regarding edge cases, missing results, or inappropriate roles.

Define four test records in advance. For 9.8 mg, 10.0 mg, 10.2 mg and a missing value, the expected outcomes are pass, pass, fail and incomplete, respectively. Check each quality decision against the defined rule. The missing value must not be treated as zero in this example.

After that, the same capture role tries to make a decision, which according to the role model is reserved for a second role. This attempt must be prevented according to the defined separation. A legitimate examiner must then see the correct results and the required history. Thus, a general acceptance becomes a concrete statement about the intended process.

Own dataset for process validation
datasetExpected process outcomeAdditional observation
L01 · 9.8 mg Passed Correct sample, batch and unit
L02 · 10.0 mg Included in accordance with the inclusive rule No change in decision due to representation
L03 · 10.2 mg Failed Further proposed processing
L04 · Value is missing Incomplete No unaudited passing as passed
Role without decision authority Decision prevents Check access and actual impact separately

Sources: [1]

End-to-end also includes transfers and errors

Check not only individual masks. If a result is transferred to an ERP, target object, unit, status and error handling are included in the statement. A successful sending does not yet prove that the target system has processed the correct dataset. The professional confirmation must reach the required decision point.

Intentionally interrupt a relevant transition in an appropriate test environment. What does the user see? Is an unfinished task detectable? Does repetition result in a duplicate process? Define the expected behavior for these scenarios prior to testing initiation. Technical error messages and subject-matter usability should be evaluated separately.

Consider the actual role hierarchy. A test conducted by an administrator may mask necessary restrictions. In contrast, representative users and handoffs between departments and QA demonstrate whether the routine workflow functions under the intended authorities.

Sources: [1]

Document test records and results in a clear and understandable manner

A robust case includes the starting condition, action, expected and actual results, and the verified status. Appendices are clearly assigned. A screenshot may serve as evidence for part of the information but does not replace an explanation of which criterion was validated. For dynamic processes, multiple observations may be required.

Record deviations even when users accept the workflow overall. A business decision to forgo a convenience feature differs from a critical data control failing its requirement. Base the response on the actual impact.

When additional criteria are introduced, it is transparent which criteria were originally defined and which have been modified subsequently. Acceptance criteria are not adjusted without notice to an unexpected result. A justified change may be possible, but it requires a clear evaluation and, if necessary, further review.

Sources: [1] [2]

Decision on takeover and release to be kept separate

The evaluation of a UAT validation certificate is a component of the final decision. Additionally, configuration, migration, training, operating procedures, and outstanding findings may be relevant. A passed process case does not automatically address all questions regarding productive usage.

Summarize the adopted UAT statements, additions, and remaining limitations. Specify the released application with sufficient detail so that the team knows which roles, process variants, and interfaces are included. Any future expansion will require a separate impact assessment.

The appropriate approach is often a shared test plan with clear acceptance criteria. User acceptance and GxP validation evidence can be derived from the same execution, provided it is properly designed and prepared. This ensures effort is focused on the required conclusions, rather than generating duplicate tests simply due to different departmental designations.

Sources: [1] [2]

Project experience and collaboration

Identify which GxP questions the user acceptance test answers

With users and QA, we inspect an executed UAT case and link its requirement, data, acceptance criterion and actual result. If the case covers the required routine process on the relevant version, its evidence may be reused. Where critical handovers or traceable results are missing, we add targeted checks. Adding a PQ heading does not create evidence. Your team should be able to explain which part of the intended use each reused or additional case demonstrates.

Frequently asked questions about UAT and PQ

Is a UAT usable without previously defined expected results?

It may provide indications regarding usability or observed behavior. However, for a specific GxP assessment conclusion, the agreed-upon criteria may still be missing. Evaluate the available evidence and appropriately supplement what is not sufficiently demonstrated. Later reclassification as PQ does not close this gap.

Sources: [1]

Must PQ tests be performed using actual production data?

Not as a general rule. Test data and conditions must represent the intended use adequately. Controlled test data may produce critical cases more reliably than random production examples. Justify the selection and include relevant volumes, variants and handovers.

Sources: [2]

Can an administrator test all process cases?

Administrative rights may obscure access restrictions and role separation. Use representative roles for these statements. An administrator may perform technical preparations; however, the actual testing must reflect the intended authorization scenario.

Is PQ required for every software a separate document?

No. In the qualification context, Annex 15 allows combining documentation for suitability (No. 2.5) and may perform the PQ together with OQ or process validation in certain cases (No. 3.13). For computerized systems, Annex 11 does not require a general, standalone file titled "PQ". The key element is appropriate, traceable evidence of the intended use. Document boundaries and combinations are defined within the quality system, taking into account function, risk, and project requirements.

Sources: [1] [2]

Must the entire UAT be repeated following a change?

Assess which test conclusions are affected by the change. Intentional additions of new roles, data rules, or interfaces may require targeted repeat and supplementary cases. A unchanged, still adequate validation evidence does not need to be regenerated solely due to a new document date.

Sources: [1]

Does a commercial acceptance automatically constitute a GxP release?

No. Contract acceptance and professional usage decisions may be based on different criteria. For usage, the relevant overall scope, findings, and required operating conditions must be evaluated. Document the decision in accordance with your own quality system.

Sources: [1]

Sources and review date

Status: October 14, 2026. The assessment of acceptance and suitability is based on the listed original sources. Examples, test matrices, and templates are standalone reference tools; adapt them to your specific application. Targeted supplemental review on October 18, 2026: Annex 15, Sections 2.5 and 3.13 regarding the combination of qualification documentation or assessments. This rule is not presented as a general requirement for a standalone software PQ document.

  1. European Commission: EU GMP Annex 11, Computerised Systems (2011)Current version as per EudraLex inventory as of 14.10.2026. Application, risks, operation, and data lifecycle; specific testing proposals are derived independently.
  2. European Commission: EU GMP Annex 15, Qualification and Validation (2015)Qualification and validation, supplemented by the specific framework for computerized systems.

Sources checked: .

Daniel Herrmann, Daniel Herrmann Consulting

Daniel Herrmann Consulting

How we can support you

Daniel Herrmann Consulting combines professional assessment with hands-on work on the specific system. During the first meeting, we clarify your intended use, existing documentation, and the next most appropriate action step.

  • Computer system validation →

    Bring your GxP systems into validated operation, from laboratory software, MES and eQMS to ERP. We define the scope, plan risk-based tests using GAMP 5 and hand the evidence to QA for release.

Send a project enquiry Customer testimonials →