Red cell porphobilinogen deaminase is known to be an indicator of the carrier state for acute intermittent porphyria (AIP). This enzyme was assayed in three groups of individuals at least 15 years old: 105 affected individuals or obligate carriers, 234 unaffected first-degree relatives of patients, and 217 unrelated control persons. Analysis of the distribution of the control enzyme activities suggested presence of three commingled distributions. Also, the overlap between carrier-group and control-group values must be taken into account for genetic counseling of relatives whose enzyme activity lies within the overlap. A Bayesian approach is proposed to derive risks for these individuals, using the observed carrier and control distributions. The method is illustrated by deriving risks for a family from our sample.