Germ-line mutation rate has been regarded classically as a fundamental biological parameter, as it affects the prevalence of genetic disorders and the rate of evolution. Somatic mutation rate is also an important biological parameter, as it may influence the development and/or the course of acquired diseases, particularly of cancer. Estimates of this parameter have been previously obtained in few instances from dermal fibroblasts and lymphoblastoid cells. However, the methodology required has been laborious and did not lend itself to the analysis of large numbers of samples. We have previously shown that the X-linked gene PIG-A, since its product is required for glycosyl-phosphatidylinositol-anchored proteins to become surface bound, is a good sentinel gene for studying somatic mutations. We now show that by this approach we can accurately measure the proportion of PIG-A mutant peripheral blood granulocytes, which we call mutant frequency, ƒ. We found that the results are reproducible, with a variation coefficient (CV) of 45%. Repeat samples from 32 subjects also had a CV of 44%, indicating that ƒ is a relatively stable individual characteristic. From a study of 142 normal subjects we found that log ƒ is a normally distributed variable; ƒ variability spans a 80-fold range, from less than 1×10⁻⁶ to 37.5×10⁻⁶, with a median of 4.9×10⁻⁶. Unlike other techniques commonly employed in population studies, such as comet assay, this method can detect any kind of mutation, including point mutation, as long as it causes functional inactivation of PIG-A gene. Since the test is rapid and requires only a small sample of peripheral blood, this methodology will lend itself to investigating genetic factors that underlie the variation in the somatic mutation rate, as well as environmental factors that may affect it. It will be also possible to test whether ƒ is a determinant of the risk of cancer.