What is the difference between Cp and Cpk?

Cp measures potential capability — it assumes the process is perfectly centered between the specification limits. Cpk measures actual capability by accounting for how far the process mean sits from the nearest limit. When the process is centered, Cp = Cpk. When off-center, Cpk < Cp.

What Cp value is acceptable?

A Cp below 1.0 means the process spread exceeds the tolerance band and produces out-of-spec parts even when centered. Cp ≥ 1.33 is the common minimum for capable processes. Cp ≥ 1.67 is required for critical-to-quality dimensions in automotive (AIAG) and aerospace.

Why can Cpk be negative?

Cpk is negative when the process mean is outside the specification limits, meaning the process is producing out-of-spec output by default. Fix the process centering before interpreting Cp.

How is sigma (standard deviation) estimated?

In practice, sigma is estimated from control chart data using Rbar/d2 (for Xbar-R charts) or Sbar/c4 (for Xbar-S charts). Using the sample standard deviation directly from all measurements gives a long-term estimate; using control chart estimates gives short-term (within-subgroup) sigma.

What does "% of tolerance used" mean?

% of tolerance used = (6 × sigma) / (USL − LSL) × 100. It shows what fraction of the available tolerance band is consumed by the 6-sigma process spread. A value of 75% (Cp = 1.33) means the spread uses three-quarters of the band, leaving a quarter as safety margin.

Is Cp the same as Pp?

Not exactly. Cp uses short-term within-subgroup sigma (from control charts) and reflects process potential when in statistical control. Pp uses long-term overall sigma (from all data) and reflects actual historical performance. This calculator accepts whichever sigma you provide.

Cp Process Capability Ratio Calculator

Calculate Cp and Cpk process capability ratios from specification limits and standard deviation to see how well a process fits within tolerance.

Process Parameters

Upper Specification Limit (USL)

Lower Specification Limit (LSL)

Process Mean

Average of measured output values.

Process Standard Deviation (σ)

Must be greater than zero.

Enter specification limits, mean, and sigma to calculate Cp and Cpk.

Cp (Potential)

—

centered process

Cpk (Actual)

—

accounting for centering

Capability Classification

—

Process Spread vs. Tolerance Band —

6σ spread / tolerance

LSL USL

Detailed Metrics

Tolerance band (USL − LSL) —

Process spread (6σ) —

% of tolerance used —

CPU (upper capability) —

CPL (lower capability) —

Mean offset from center —

Cp / Cpk Interpretation Guide

Cp / Cpk < 1.00 Not capable — produces defects

1.00 ≤ Cp / Cpk < 1.33 Marginal — monitor closely

1.33 ≤ Cp / Cpk < 1.67 Capable — meets most standards

Cp / Cpk ≥ 1.67 Excellent — high confidence

Cpk drives the classification. A high Cp with a low Cpk signals an off-center process that needs mean adjustment.

Summary

Calculate Cp and Cpk process capability ratios from specification limits and standard deviation to see how well a process fits within tolerance.

How it works

Enter the Upper Specification Limit (USL) and Lower Specification Limit (LSL).
Enter the process mean (average measured output value).
Enter the process standard deviation (sigma).
Cp = (USL − LSL) / (6 × sigma) — measures potential capability assuming a centered process.
Cpk = min((USL − mean) / (3 × sigma), (mean − LSL) / (3 × sigma)) — measures actual capability accounting for centering.
Results show both indices, tolerance used, process spread, and a classification from not capable to excellent.

Use cases

Assess whether a machining process can hold tight dimensional tolerances.
Compare capability before and after a process improvement project.
Identify off-center processes where Cpk is much lower than Cp.
Set sigma targets to achieve a desired Cp of 1.33 or 1.67.
Support PPAP (Production Part Approval Process) submissions with Cp/Cpk data.
Train quality engineers on the difference between potential and actual capability.

Frequently Asked Questions

Last updated: 2026-09-20 · Reviewed by Nham Vu