Risperidone is metabolized by polymorphic enzymes, and a large variability in plasma concentration and therapeutic response is observed. Risperidone long-acting injection (RLAI) avoids the first-pass effect, and little is known about the influence of gene polymorphisms involved in its pharmacokinetics. The influence on plasma concentrations of risperidone (RIS), its metabolite 9-hydroxy-risperidone, and on adverse effects were investigated for polymorphisms of cytochrome P450 2D6 (CYP2D6) (*3, *4, *5, *6), CYP3A (CYP3A4*1B, CYP3A4 rs4646437, CYP3A5*3, CYP3A7*1C), ABCB1 (1236C>T, 2677G>T, 3435C>T), NR1/2 coding for pregnane X receptor (rs1523130, rs2472677, rs7643645), and for CYP3A activity measured by a phenotyping test. Forty-two patients with at least 4 consecutive unchanged doses of RLAI were included in a multicenter cross-sectional study. A 55% lower dose-adjusted plasma levels of RIS were observed for CYP2D6 ultrarapid metabolizers (n = 5) as compared with CYP2D6 intermediate metabolizers (P < 0.007). NR1/2 polymorphism (rs7643645A>G) influenced RIS exposure with a 2.8-fold lower active moiety (P = 0.031) in GG compared with the AA genotype. This was confirmed in a second independent cohort (n = 16). Furthermore, high-density lipoprotein cholesterol was positively correlated with CYP3A activity (P = 0.01), and the NR1/2 (rs2472677) polymorphism was associated with different adverse effects including prolactin plasma levels adjusted for age and sex. In conclusion, our results confirmed the influence of CYP2D6 genotype on plasma levels of RIS. This is the first report on the influence of NR1/2 polymorphisms on RLAI exposure and on drug-induced adverse effects. These results should be validated in larger cohorts.