KaizumiKaizumi™

Cp / Cpk

Cp, Cpk & Capability Analysis

1/6

What is Process Capability?

Process capability compares the voice of the process (how much your output naturally varies) against the voice of the customer (your specification limits). It answers one question: can this process reliably produce output inside spec? The two headline indices are Cp (potential capability) and Cpk (actual capability, which also accounts for how centered the process is).

Capability only means something for a stable, in-control process. If your process is still drifting or has special-cause variation, fix that first — a capability index on an unstable process is a snapshot of chaos, not a prediction.
Record Info (optional)
Load an Industry Example

Capable, marginal, and not-capable processes — see how centering and spread drive Cpk.

Spec Limits & Process Statistics

Leave blank if one-sided

Leave blank if one-sided

Average of your data

Process sigma (σ)

Cpk
Enter at least one spec limit, the mean, and the standard deviation
Or pick an industry example above

Want an editable copy?

Download the free Cp/Cpk template — enter spec limits, mean, and standard deviation and it calculates Cp, Cpk, estimated PPM out of spec, and the sigma level, with capability bands shaded automatically. Opens in Excel and Google Sheets.

What is process capability?

Process capability compares the voice of the process (how much your output naturally varies) against the voice of the customer (your specification limits). The two headline indices are Cp, which asks whether the process spread could fit inside the spec width, and Cpk, which also accounts for how well-centered the process is. A capability study is a standard step in Six Sigma and in production part approval (PPAP).

Capability indices only mean something for a stable, in-control process. If the process is still drifting or shows special-cause variation, address that first — an index computed on an unstable process describes chaos, not future performance.

Cp, Cpk, and the sigma level

Capability indices and sigma levels are two languages for the same margin between your process and the edge of spec. Cpk 1.33 ≈ 4 sigma, 1.67 ≈ 5 sigma, and 2.00 ≈ 6 sigma. This tool reports the short-term sigma (Zst = 3 × Cpk) alongside Cp and Cpk, and estimates the parts-per-million out of specification directly from your process statistics.

When to use this tool

Run a capability study before releasing a new process, as part of PPAP, or to prove an improvement. If Cpk is low, the curve tells you why: a wide curve is a variation problem, while a curve pushed toward one limit is a centering problem — usually the cheaper fix. To find the root causes of the defects driving low capability, pair this with a Fishbone diagram and a Pareto chart.

Frequently asked questions

What is the difference between Cp and Cpk?
Cp measures potential capability — whether the process spread could fit inside the spec width — using Cp = (USL − LSL) ÷ (6 × standard deviation). It ignores where the process is centered. Cpk measures actual capability and is the smaller of (USL − Mean) ÷ (3σ) and (Mean − LSL) ÷ (3σ), so it penalizes an off-center process. Cpk is always less than or equal to Cp; they are equal only when the process is perfectly centered between the limits.
What is a good Cpk value?
A Cpk of 1.33 is the common industry minimum for a capable process — about 63 defective parts per million and roughly a 4 sigma process. 1.00 to 1.33 is marginal, and below 1.00 the process cannot reliably meet specification. Many automotive and aerospace programs require 1.67 (about 5 sigma) for critical characteristics, and 2.00 corresponds to a 6 sigma process.
How do you calculate Cpk?
Cpk = min[(USL − Mean) ÷ (3 × standard deviation), (Mean − LSL) ÷ (3 × standard deviation)]. Compute both one-sided indices — Cpu toward the upper limit and Cpl toward the lower limit — and take the smaller. For a one-sided specification, use whichever limit exists. This calculator does it automatically and also estimates the resulting parts-per-million out of spec.
What is the difference between Cpk and Ppk?
Cpk uses the within-subgroup (short-term) standard deviation and describes the capability the process could achieve when in control. Ppk uses the overall (long-term) standard deviation from all the data and describes actual performance including drift between subgroups. The formulas are identical; only the standard deviation differs. When you estimate sigma from all your data at once — as in this tool — the result is closest to Ppk, and Cpk ≈ Ppk when the process is stable.
Can I calculate Cpk with a one-sided specification?
Yes. Many characteristics have only an upper limit (for example contamination or turnaround time) or only a lower limit (for example strength). Leave the unused spec limit blank and the tool computes the one-sided index against the limit that exists. Cp is not defined for a one-sided spec because there is no spec width, so only Cpk is reported.
Is this Cpk calculator free?
Yes. It runs in your browser with no sign-up, accepts either summary statistics or pasted raw data, draws a capability curve with your spec limits, and exports a print-ready study. A free editable Excel/Google Sheets template is also available.

Related tools