Process capability indices have been introduced for measuring process reproduction capability of a manufacturing industry. The loss-based capability index $$C_{pm}$$ takes into account the degree of process targeting (centering), which essentially measures the process performance on the basis of average process loss. Generally, the underlying manufacturing process data are obtained from the output responses of continuous quantities, which are always assumed to be real numbers. However, in a practical situation, the data collected by output process measurements are often imprecise (fuzzy data). We propose a constructive methodology for obtaining the fuzzy estimate of $$C_{pm}$$ using fuzzy data, which is based on “resolution identity” in fuzzy sets theory. We propose four decision rules to judge the process condition by simultaneously introducing randomness and fuzziness. Finally, we present a sequence of testing steps for assessing manufacturing process performance using the critical value of $$C_{pm}$$ with fuzzy data and can be easily implemented in real situations.