Cpk (process capability index) measures how well a process fits within its specification limits, accounting for how off-center it is. It is the smaller of (USL − mean)/(3σ) and (mean − LSL)/(3σ). Higher is better.

What is a good Cpk value?

A Cpk of 1.33 is the common industry threshold for a capable process (~63 DPMO). 1.00 means the process barely fits its spec; 1.67+ is often required for critical or safety characteristics.

What is the difference between Cpk and Ppk?

Cpk uses within-subgroup (short-term) sigma, estimated here from the average moving range (R̄/d₂). Ppk uses overall (long-term) sigma, the ordinary sample standard deviation. Cp/Cpk describe potential; Pp/Ppk describe actual performance.

How many data points do I need?

You need at least 2 to compute anything, but for a trustworthy capability estimate use 30+ measurements from a stable, in-control process. Capability on an unstable process is not meaningful.

Free tool, no signup

Cpk calculator

Paste your process data and spec limits. You get Cp, Cpk, Pp, Ppk, your sigma level and DPMO, computed in your browser. Nothing is uploaded.

Process data

Lower Spec Limit (LSL)

Upper Spec Limit (USL)

Runs in your browser. Nothing is uploaded. Same engine as SigmaForge DataForge™.

Enter your data and spec limits, then hit Calculate to see Cp, Cpk, Pp, Ppk, your sigma level, and DPMO.

How Cpk is calculated

Process capability compares the voice of the process, its natural variation, to the voice of the customer, the specification limits. This calculator does it the way a Green Belt would: it estimates the mean and both the short-term (within) and long-term (overall) standard deviation from your data, then derives Cp, Cpk, Pp and Ppk.

Cpk = min( (USL − x̄) / 3σ , (x̄ − LSL) / 3σ )

What the number means

Cpk	Verdict	Approx. DPMO
< 1.00	Not capable	> 2,700
1.00 to 1.33	Marginal	63 to 2,700
1.33 to 1.67	Capable	0.6 to 63
1.67 and above	Excellent	< 0.6

Questions people ask

What is Cpk?: Cpk (process capability index) measures how well a process fits within its specification limits, accounting for how off-center it is. It is the smaller of (USL − mean)/(3σ) and (mean − LSL)/(3σ). Higher is better.
What is a good Cpk value?: A Cpk of 1.33 is the common industry threshold for a capable process (~63 DPMO). 1.00 means the process barely fits its spec; 1.67+ is often required for critical or safety characteristics.
What is the difference between Cpk and Ppk?: Cpk uses within-subgroup (short-term) sigma, estimated here from the average moving range (R̄/d₂). Ppk uses overall (long-term) sigma, the ordinary sample standard deviation. Cp/Cpk describe potential; Pp/Ppk describe actual performance.
How many data points do I need?: You need at least 2 to compute anything, but for a trustworthy capability estimate use 30+ measurements from a stable, in-control process. Capability on an unstable process is not meaningful.

Go beyond the number

DataForge adds the capability histogram, a written interpretation and a branded PDF, and it is one of 32 analyses. White Belt is free, including the exam and the certificate.

Create a free account

How Cpk is calculated

What the number means

Questions people ask

Go beyond the number

More in resources