Yield rate is the percentage of units that pass all inspection and test steps in a manufacturing process without rework or scrap. In medical devices, it is measured per process step or across a full line, and it feeds process validation, statistical process control, and nonconformance reporting under ISO 13485:2016.
What is the yield rate?
Yield rate compares the good output against what was started. A line releasing 940 conforming units from 1,000 started boards has a 94 percent yield. The number is simple. What it hides is not.
Three variants matter. First pass yield (FPY) counts only units that passed every step correctly the first time, with no rework. Final yield counts everything that eventually shipped, including reworked units. Rolled throughput yield (RTY) multiplies the FPY of each step, so a ten-step process running 99 percent per step lands near 90 percent overall. Yield rate sits downstream of design transfer, inside production and process controls, and it is one of the earliest signals that a validated process has drifted.
Why does yield rate matter in medical device development
Scrap is expensive in any industry. In a regulated one, the cost is regulatory as well as financial.
Every failed unit generates a nonconformance record, and repeat nonconformances against one defect code are a standard CAPA trigger. An auditor reading a yield trend that fell from 96 to 82 percent over two quarters, with no corresponding investigation, has found a finding.
The FDA’s Quality Management System Regulation (21 CFR Part 820), effective February 2, 2026, incorporates ISO 13485:2016 by reference. That makes clause 8.4, analysis of data, and clause 8.3, control of nonconforming product, the live citations. Yield data is exactly what those clauses expect you to analyze. Under EU MDR 2017/745, the same trend feeds quality management system obligations.
A patient safety argument sits underneath the paperwork. Yield loss on a critical to quality (CtQ) characteristic means the process is producing units near the specification limit. Escape rate rises with yield loss, and for a Class III implant, the escaping unit is the one that matters.
How the yield rate is calculated and controlled
The calculations are straightforward:
- First pass yield = units passing all steps with no rework / units started.
- Final yield = conforming units released / units started.
- Rolled throughput yield = FPY(step 1) × FPY(step 2) × … × FPY(step n).
Control comes from four places. Statistical process control charts are on the parameters that drive the defect, not on the defect count itself, because a chart of failures tells you the process broke after it broke. Process capability indices (Cp and Cpk) on CtQ dimensions, with Cpk of 1.33 as a common internal floor for safety-related characteristics. Process validation under GHTF/SG3/N99-10:2004, the guidance the FDA recognizes for device process validation, where operational qualification establishes the parameter window that yields a conforming product. And traceability, usually through a manufacturing execution system, so an excursion resolves to a lot, a shift, or a supplier batch.
PFMEA links the two halves. It predicts where yield will be lost; the yield data tells you whether the prediction was right.
Common challenges and best practices
The most common failure is reporting final yield and never computing FPY. Rework absorbs the signal. A line at 98 percent final yield and 71 percent first pass yield is a line with a very busy rework station, and rework on a validated process carries its own requirement: ISO 13485:2016 clause 8.3.4 expects rework instructions to be documented and the adverse effect of rework on the product to be assessed.
Second is inspection pressure. When yield becomes an operator metric, marginal units start passing. Gauge repeatability and reproducibility studies exist partly to catch this. If measurement variation is a large share of tolerance, your yield number is measuring the gauge, not the process.
Third is low volume. Class III lines often run lots of 30 units. Yield percentages over small lots are noisy, and teams overreact to single-lot swings. Trend across lots and set investigation thresholds on cumulative counts, not percentages.
What good looks like: FPY tracked per step, RTY per line, defect codes shared between the yield and nonconformance systems, capability studies refreshed after every change control touching a validated parameter, and a yield threshold that escalates to root cause analysis before it escalates to CAPA.
How SJML helps with yield rate
Syrma Johari MedTech runs medical device manufacturing across ISO Class 7 and Class 8 cleanrooms and ESD-controlled environments, with high-speed SMT lines instrumented for solder paste inspection, automated optical inspection, and X-ray inspection. Those inspection points are where first-pass yield data is generated. Process validation (IQ, OQ, PQ), PFMEA, PPAP, and SAP-integrated MES traceability tie yield excursions back to lot, station, and supplier. Supplier qualification and dual sourcing address the incoming material variation that drives much of the loss. Nonconformance handling, root cause analysis, and CAPA close the loop.
Talk to SJML’s manufacturing team →
Frequently asked questions
There is no universal target. Yield depends on process complexity, tolerance tightness, and volume. A mature SMT line may exceed 98 percent first-pass yield, while a new molded component with tight geometric tolerances may start near 80 percent and climb through optimization. The useful benchmark is your own validated baseline, and any sustained drop below it warrants investigation.
First pass yield measures a single step: units that passed without rework, divided by units entering. Rolled throughput yield multiplies the first-pass yield of every step, giving the probability that a unit passes the entire process untouched. RTY is always lower than any individual FPY, and it exposes cumulative loss that step-level reporting hides.
Not automatically. A yield drop is data, not a nonconformity by itself, and it should trigger an investigation under your data analysis procedure. If root cause analysis shows a systemic process failure, or the same defect recurs across lots, CAPA criteria are typically met. Document the decision either way, since auditors ask why a visible trend produced no action.
No regulation names yield rate directly. ISO 13485:2016, incorporated by reference into the FDA QMSR, requires monitoring and measurement of processes and analysis of quality data. Yield is the ordinary way manufacturers meet that expectation for production. EU MDR 2017/745 imposes equivalent quality management system obligations. The metric is not mandated; demonstrating process control is.
Related terms
- Statistical Process Control (SPC)
- Process Validation
- Nonconformance Report (NCR)
- Critical to Quality (CtQ)
- Corrective and Preventive Action (CAPA)
