First pass yield (FPY) is the percentage of parts that come out of an operation good on the first try, with no scrap, rework or retest. You calculate it by dividing first-pass good parts by the parts you started, then multiplying by 100. If 1,000 parts go in and 950 come out good the first time, your FPY is 95%.
The formula takes ten seconds. The hard part is deciding what counts as "good the first time" and then using the number to find where your shop loses time and material. This guide covers both, with worked CNC examples.
TL;DR
- FPY = first-pass good parts ÷ parts started × 100. Rework and retests do not count as first-pass good.
- Measure FPY per operation, not just per job. Multiply the operations together to get rolled throughput yield (RTY), which shows the true flow-through rate.
- A reported median across many industries is 95%, but no single percentage is a target for every process. Compare an operation with its own history first.
- Low FPY costs machine time as well as material. Every reworked part takes spindle hours the schedule did not plan for.
- Find the operation with the worst FPY, then fix its cutting parameters, offsets or tool wear.
Table of Contents
- What first pass yield measures
- The first pass yield formula
- How to count rework, retests and repairs
- First pass yield versus rolled throughput yield
- What is a good first pass yield?
- How FPY changes your cycle time and capacity
- Using FPY to find and cut scrap
- Our perspective
- Put the numbers to work with Availzye Machinist Pro
- A simple FPY routine for your shop
- FAQ
- Sources
What first pass yield measures
The ASQ glossary gives the standard definition. FPY, also called the quality rate, is the percentage of units that complete a process and meet quality guidelines without being scrapped, rerun, retested, returned or diverted to an offline repair area. Minitab's documentation puts it as the share of product that is acceptable the first time it goes through the process.
For a machine shop, that means a part counts only if it comes off the machine in tolerance, with the right finish, and goes on to the next step without a second pass. A part that gets touched up on the bench, re-cut on a second setup, or re-inspected after an adjustment is not first-pass good, even if it ships.
That distinction is the reason FPY is useful. A plain final yield asks whether you could ship the part. FPY asks what it cost you to get there.
The first pass yield formula
The formula appears in several forms, but they all say the same thing. The common shop version is:
FPY (%) = (first-pass good parts ÷ total parts started) × 100
LNS Research writes it as units completed to specification with no rework, divided by total units entering the process. Their example: 100 units enter, 99 are finished to specification and 2 are reworked, so FPY = (99 − 2) ÷ 100 = 97%. The reworked pieces were finished, but they come out of the numerator.
Minitab's equivalent is 100% minus the defect percentage, which gives the same answer when "defect" means any part that did not pass the first time.
A CNC worked example
A CNC yield guide from PrecisionAM uses a batch of 1,000 parts started. Of those, 950 pass first time, 30 need rework and 20 are scrapped. The three numbers that matter:
| Metric | Formula | Result |
|---|---|---|
| First pass yield | 950 ÷ 1,000 × 100 | 95% |
| Scrap rate | 20 ÷ 1,000 × 100 | 2% |
| Final yield | (950 + 30) ÷ 1,000 × 100 | 98% |
The final yield of 98% looks healthy. The FPY of 95% tells you that 5 parts in every 100 needed a second look, and 3 of those 5 were saved only by extra work.
The same source gives the formula for ordering stock. To get the quantity of blanks needed, divide the good parts required by one minus the scrap rate: Q₀ = Q ÷ (1 − q). For 1,000 good parts at a 4% scrap rate, that is 1,000 ÷ 0.96, so you start 1,042 blanks.
How to count rework, retests and repairs
Most FPY arguments in a shop come down to one question: does a reworked part count? Under the ASQ definition above, it does not. A unit that is scrapped, rerun, retested or sent to repair has failed the first pass.
Before you start tracking, write down the rules and apply them the same way every shift:
- What is one unit? One part, one fixture plate of parts, or one lot? Count individual parts if you can.
- Where does the process start and stop? FPY belongs to a defined operation, such as roughing, finish boring or deburr.
- What counts as a pass? Use the print: dimensions, surface finish, thread fit and cosmetic requirements.
- What counts as rework? Any second cut, hand fitting, re-measurement after an offset change, or re-inspection.
- Does a first-article adjustment count? Decide this once. Many shops exclude the setup piece from the count, but only if they say so and do it every time.
Write these answers into the work order or the shop's quality procedure so a night-shift operator and a day-shift lead record the same thing.
First pass yield versus rolled throughput yield
FPY is a per-operation number. A part that goes through saw, rough mill, finish mill, drill and tap, deburr and inspect has six chances to fail, and a single job-level FPY hides which one did it.
Rolled throughput yield (RTY) multiplies the FPY of every operation to give the chance that a unit gets through the whole sequence first time. Larry Fast's IndustryWeek piece makes the point with numbers: six operations each reporting 95% FPY give an RTY of only 74%. You can check it yourself: 0.95 to the sixth power is about 0.735.
That is why a shop where every operation "looks fine" can still feel like it is always chasing parts. The quick comparison:
| FPY | RTY | |
|---|---|---|
| Scope | One operation | The whole routing |
| Calculation | Good first time ÷ started | Product of each operation's FPY |
| Use it to | Find the problem operation | See the real flow-through of a job |
Use FPY to locate the problem and RTY to understand how much it costs the job.
What is a good first pass yield?
There is no single answer, and you should be wary of anyone who gives you one. Still, benchmarks help you place yourself.
- APQC's open benchmarking data reports a median finished-product first-pass quality yield of 95.0% across 5,453 companies. That is one measure for one population, from many industries, and not a machining standard.
- For precision machining specifically, PrecisionAM reports scrap of 1 to 3% as typical for mid-market shops, above 5% as a sign of weak process control, and below 0.5% as world class. It puts typical aerospace FPY at roughly 90 to 94%, with world-class flight-hardware lines at 97 to 99%. These are one company's published ranges, so treat them as orientation.
Three things matter more than any benchmark:
- Part complexity. Thin-walled aerospace features, tight-tolerance bores and long-reach finishing passes will never yield like a simple turned pin.
- Batch size. In high-mix, low-volume work you rarely get long runs to perfect a process, so first-article risk weighs more heavily.
- Trend. An operation moving from 91% to 95% over a quarter has improved more than one sitting at 96% with no plan.
If you want more context for reading numbers like these against your own equipment, a practical way to start is to log counts straight from the controller. If your cells run Allen-Bradley controllers, the Studio 5000 produced and consumed tag setup walkthrough shows how controllers share data, which makes it easier to pull consistent counts instead of relying on tally sheets.
How FPY changes your cycle time and capacity
Cycle time per part is not the same as machine time per good part. When FPY drops, every good part carries a share of the time spent on parts that failed.
Here is the simple math, with illustrative numbers. Say a part has a 6-minute cycle and the operation runs at 93% FPY. Ignoring rework time, you spend 6 ÷ 0.93 = 6.45 minutes of spindle time for every first-pass good part. At 97% it is 6.19 minutes. Across a 500-part order, that gap is about 2 hours of machine time you could sell elsewhere.

Cost works the same way. Material and machine cost per part, divided by one minus the scrap fraction, gives the cost per good part. At 4% scrap, a part that costs $20 to make costs about $20.83 per good part. Our guide to machining cost breakdown shows where each of those dollars lands.
Rework is the hidden part. PrecisionAM gives an example: if each of 40 rework attempts takes 12 more minutes on the same machining center, rework consumes 480 minutes, or eight machine hours, even when the final yield report looks acceptable. That is a full shift that was never in the schedule.
FPY also feeds overall equipment effectiveness. The ASQ glossary notes that OEE is the product of availability, performance efficiency and first-pass yield. LNS Research adds that FPY on its own does not capture downtime, schedule or capacity, so track it next to those measures. If you estimate quotes, feed your real FPY into the estimate, as covered in our guide to machining cost estimation.
Using FPY to find and cut scrap
A yield number only pays off if it points to a fix. The approach that works in practice:
- Track FPY by operation. Roughing, semi-finish, finish and secondary ops each get their own number.
- Sort rejects by cause. Dimensional, surface finish, tool breakage, fixture, programming.
- Fix the worst operation first. One operation usually accounts for most losses.
- Re-measure for a full cycle of shifts before declaring success.
Two published CNC examples show what this looks like. Both come from a machining service company, and the figures are its own reported results, so use them as illustrations rather than promises.
In the first, an aerospace supplier found that 68% of its dimensional rejects originated in roughing. Changing the depth of cut from 0.120″ to 0.090″ and raising feed by 15% moved roughing FPY from 79% to 94% and cut overall scrap by 41% in the first month. The lesson: the loss was upstream of where the parts were being caught.

In the second, a Michigan transmission housing line ran 720 pieces per shift on three vertical machining centers at 93% FPY. Night shift dropped to 87% because a probe offset was re-entered by hand after every tool change. A small macro that stored the offset in a persistent variable lifted FPY to 95% within a week. The lesson: look at what differs between shifts before blaming the operators.
Common causes worth checking
- Cutting parameters. Feeds and speeds set too aggressive for the tool and material produce chatter, poor finish and out-of-tolerance walls. See our guide to calculating feeds and speeds.
- Tool deflection. Long-reach tools push off the cut and leave tapered walls. Our piece on a tool deflection calculator explains how to estimate it before the first part.
- Spindle load. Overloaded cuts lead to stalls, tool wear and drift. Check the numbers with the spindle power formula.
- Surface finish. Scallops left by stepover choices fail finish callouts. See scallop height.
- Machine condition. Wear in ways, spindles and coolant systems shows up as drift. A CNC maintenance checklist tied to run hours helps catch it before parts fail.
- Manual steps. Offsets, work coordinates and tool length entries typed by hand are a repeat offender.
Anebon's write-up suggests pulling data from the controller itself: the M30 count for completed parts, the alarm log for stops, and the tool-life counter for inserts used. Plotting those on a control chart shows process drift before it becomes scrap.
Our perspective
Most shops we talk to already know their scrap pile. Fewer know where it started. FPY is the number that closes that gap, but only if the inputs live in one place.
In our experience the biggest losses hide in the handoffs: a feed rate chosen by feel, a tool that has run past its useful life, a work order that sits on paper while the part is already in the vise. Each of those is a small decision. Together they decide your yield.
That is why we built Availzye Machinist Pro as a single cloud app. The calculation tools for feeds and speeds, tool deflection, power and torque, chip thinning and chatter prediction help set a cutting strategy before the first chip. The Job Tracker, Tool Crib with tool life logging, and Maintenance Tracker keep the shop-floor context next to those numbers, so when a part fails you can ask what changed instead of guessing.
Our aim is simple: fewer surprises on the first part, and less switching between spreadsheets, paper logs and separate apps to find out why.
Availzye
Put the numbers to work with Availzye Machinist Pro
If you run real jobs and want parameters, tool tracking and work orders in one place, Availzye Machinist Pro is built for that. It combines machining calculators with shop management features, including a Job Tracker, Maintenance Tracker, Tool Crib inventory with low-stock alerts and tool life logging, and role-based team access.

There are three plans: Individual at $9.99 CAD per month, Small Shop at $24.99 CAD per month and Team at $49.99 CAD per month, each with a 7-day free trial. You can compare them on the Availzye Machinist Pro plans page, or go straight to the feeds and speeds calculator or the tool deflection calculator to check a tough setup before cutting.
A simple FPY routine for your shop
You do not need a new system to start. Here is a routine you can run this month:
- Pick your three highest-volume or highest-scrap parts.
- For each, list the operations from stock to ship.
- Write the pass rules for each operation (see the rework rules above).
- Count started, first-pass good, reworked and scrapped for each operation across one full month.
- Calculate FPY per operation and RTY for the routing.
- Pick the lowest operation and run one change at a time: feed, depth of cut, tool, offset routine or fixture.
- Recalculate after enough parts to see a real difference, not after one good shift.
Keep the monthly numbers. A single month tells you where you stand; six months tell you whether you are improving. Use the trend, plus your own part mix, to set targets for each operation, instead of borrowing a percentage from a benchmark.
FAQ
What is the formula for first pass yield?
FPY equals first-pass good parts divided by total parts started, multiplied by 100. If 100 parts start and 90 come off good with no rework, FPY is 90%. Rework, retests and scrap all count against the first pass.
Does a reworked part count toward FPY?
No. Under the ASQ definition, a unit that is scrapped, rerun, retested or sent to repair is not first-pass good. The part may still ship and show up in final yield, but it stays out of the FPY numerator.
What is a good first pass yield for a CNC shop?
It depends on part complexity, tolerances and batch size. APQC's cross-industry median is 95.0%, and published aerospace machining ranges run from about 90 to 94% typical up to 97 to 99% for world-class lines. Track your own trend by operation and improve from there.
How is FPY different from rolled throughput yield?
FPY measures one operation. RTY multiplies the FPY of every operation in the routing to show the chance a part passes everything first time. Six operations at 95% each give an RTY of about 74%.
Sources
- ASQ Quality Glossary
- Minitab Workspace: throughput yield percent
- LNS Research: First Pass Yield Benchmark Data
- PrecisionAM: CNC Machining Yield Rates
- IndustryWeek: More Must Have Metrics
- APQC: Finished-product first-pass quality yield
- Anebon: CNC machining scrap reduction
- Anebon: CNC machining production yield
- Studio 5000 produced and consumed tag setup
