The objective of this study was to analyze the relationship between CYP2D6 genotype and pharmacokinetics and pharmacodynamics of risperidone. Seventy-one healthy volunteers (36 women and 35 men) received a 1-mg single oral dose of risperidone. Six major CYP2D6 polymorphisms (CYP2D6*3, *4, *5, *6, *7, and *9) and the duplication were detected. Subjects were classified into 4 phenotypic groups: 6 ultrarapid (UMs), 34 extensive (EMs), 25 intermediate (IMs), and 6 poor metabolizers (PMs). There was a clear relationship between the number of active alleles and the pharmacokinetic parameters for risperidone and 9-hydroxyrisperidone, but there were no differences for total active moiety. Area under the curve and half-life of risperidone were significantly higher in PMs and IMs compared with EMs and UMs, which showed higher area under the curve of 9-hydroxyrisperidone. Risperidone produced a small decrease in blood pressure, a mild increase in QTc and a quick increase in prolactin, without significant differences between groups. Surprisingly, the incidence of adverse reactions was lower in PMs (50%) than in other subjects (78%). In conclusion, metabolism of risperidone depends on the number of active CYP2D6 alleles. So, PM subjects show higher concentrations of risperidone and very low concentrations of 9-hydroxyrisperidone. On the contrary, EM and UM subjects show low concentrations of risperidone and high concentrations of 9-hydroxyrisperidone. However, no major pharmacodynamic differences are observed between CYP2D6 genotypes, presumably because of the similar pharmacological activity of parent drug and metabolite.