How many runs does a process validation need?

The number your protocol and risk assessment require; the runs must be consecutive and reproduce the process. This page multiplies the per run sample by the run count you enter, so a five run plan pulls more samples than a three run plan.

Why does the minimum override the sampling rate?

Because a protocol floor exists to keep the count valid when a percentage of a small run would fall short. When the calculated share sits below the floor, the minimum governs and lowering the rate further saves no samples.

Does the calculator round sample counts up?

Yes. A fractional sample becomes the next whole device, and the page then takes the larger of that count and the minimum. Enter whole units per run so the rounding tracks the devices actually built.

Can I size an attribute sampling plan with this page?

For the count, yes. The acceptance number, confidence and reliability targets belong in the statistical rationale. NIST's acceptance sampling chapter explains how lot size, acceptance number and sample size interact for attribute plans.

MedTech Manufacturing calculator

Process Validation Sample Size Calculator

Size the sample plan for process validation before the protocol is approved. Enter the runs, units per run, sampling rate and minimum to see whether the rate or the floor decides the count.

What this calculator does

Size the sample count for process validation runs and see whether the sampling rate or the protocol minimum governs it.

Formula used

Samples per run = MAX(minimum samples per run, CEILING(units per run × sampling rate ÷ 100, 1))
Total validation samples = samples per run × validation runs
Total under minimum rule = minimum samples per run × validation runs

Inputs explained

Validation Runs: Consecutive PQ runs the protocol covers.
Units per Run: Devices built in each validation run.
Sampling Rate per Run: Share of a run's units pulled for testing.
Minimum Samples per Run: Protocol floor for statistical validity per run.

How to use the result

Best suited to sizing samples for a three run validation, checking whether a sampling floor binds, budgeting PQ test samples before approval.
The page sizes the count, not the statistical rationale: confidence, reliability and acceptance criteria stay in the protocol. It covers non-destructive sampling plans; destructive tests need extra units budgeted separately.

Current U.S. benchmarks

U.S. manufacturing runs at 75.7% of capacity with new factory orders at $663B per month (Federal Reserve and Census, Aug 2026).
The U.S. has 8,825 medical equipment and supplies establishments employing about 308,388 workers (Census County Business Patterns, 2023).

Common questions

How many runs does a process validation need? The number your protocol and risk assessment require; the runs must be consecutive and reproduce the process. This page multiplies the per run sample by the run count you enter, so a five run plan pulls more samples than a three run plan.
Why does the minimum override the sampling rate? Because a protocol floor exists to keep the count valid when a percentage of a small run would fall short. When the calculated share sits below the floor, the minimum governs and lowering the rate further saves no samples.
Does the calculator round sample counts up? Yes. A fractional sample becomes the next whole device, and the page then takes the larger of that count and the minimum. Enter whole units per run so the rounding tracks the devices actually built.
Can I size an attribute sampling plan with this page? For the count, yes. The acceptance number, confidence and reliability targets belong in the statistical rationale. NIST's acceptance sampling chapter explains how lot size, acceptance number and sample size interact for attribute plans.

Last reviewed 2026-10-01.