Metabolic activity of the polymorphicN-acetyltransferase (NAT2) is determined by the mutation pattern of theNAT2gene. This results in interindividual differences in the metabolic capacity (the phenotype), with continuous distribution of the activities rather than qualitative distinction between rapid and slow acetylators. To determine whether the phenotype might be predicted solely from the mutation pattern ofNAT2, quantitative relationships were calculated between mutation patterns of theNAT2gene and the phenotype of NAT2 assessed either in vitro or in vivo. Healthy volunteers were examined for the velocity at which they metabolized sulfamethazine, and human liver cytosols were measured for NAT2 enzymatic activity, obtaining in vivo and in vitro metabolic phenotype, respectively. Typing of theNAT2gene was performed by polymerase chain reaction, restriction fragment length analysis, or allele-specific polymerase chain reaction. Multiple linear regression analysis provided quantitative relationships between allelic pattern and the NAT2 activities measured in vivo and in vitro. Estimates showed the influence of particular allelic configurations on enzyme activity in vitro and the extent of acetylation of the probe drug in vivo, resulting in a strict gene–dose effect. Comparison of in vitro results with in vivo phenotyping figures showed a high degree of correspondence, indicating that the one is the reflection of the other.