AbstractTramadol is an important minor opioid prescribed for pain management. In this study, we analyzed the well‐known impact of CYP2D6 genetic variation and 60 additional variants in eight candidate genes (i.e., ABCG2, SLCO1B1, CYP2D6, CYP2B6, CYP2C19, CYP2C9, CYP3A5, and CYP3A4) on tramadol efficacy and safety. Some 108 patients with pain after surgery admitted to a post‐anesthesia care unit (PACU) and prescribed tramadol were recruited. They were genotyped, and tramadol M1/M2 metabolite concentrations were determined by a newly validated HPLC‐MS/MS method. CYP2D6 intermediate (IM) and poor (PM) metabolizers showed lower M1 concentrations adjusted for dose/weight at 30 and 120 min compared to ultrarapid (UM) and normal (NM) metabolizers (univariate p < 0.001 and 0.020, multivariate p < 0.001 and 0.001, unstandardized β coefficients = 0.386 and 0.346, R² = 0.146 and 0.120, respectively). CYP2B6 PMs (n = 10) were significantly related to a higher reduction in pain 30 min after tramadol intake (univariate p = 0.038, multivariate p = 0.016, unstandardized β coefficient = 0.224, R² = 0.178), to lower PACU admission time (p = 0.007), and to lower incidence of adverse drug reactions (p = 0.038) compared to the other phenotypes. CYP3A4 IMs and PMs showed a higher prevalence of drowsiness and dizziness (p = 0.028 and 0.005, respectively). Our results suggest that the interaction of CYP2B6 and CYP2D6 phenotypes may be clinically relevant, pending validation of these results in large, independent cohorts. Additional research is required to clarify the impact of CYP3A4 genetic variation on tramadol response.