Therapeutic drug monitoring (TDM) data for the antipsychotic drug olanzapine were investigated with respect to concentration versus dose relationship, intraindividual versus interindividual variability, and the combined influence of patient characteristics on steady-state concentration. The study included 250 patients, with daily doses ranging from 2.5 to 30 mg. Median concentration to dose ratio was 2.1 (ng/mL)/(mg/d), with 90% of the distribution in a fivefold range. In the first subgroup of patients with two measurements at different doses (n = 21), data were in keeping with linear concentration versus dose relationship. In the second subgroup of patients with repeated measurements at a constant daily dose (n = 40), estimates of within-patient and between-patient variabilities were 30.5% and 49.4%, respectively. In the whole sample, multiple regression analysis of dose-normalized concentration revealed significant effects of time postdose (−18% per 12 hours delay, P < 0.05), age ≥60 years (+27%, P < 0.005), cigarette smoking (−12%, P < 0.05), and comedication with fluvoxamine (+74%, P < 0.001), paroxetine, fluoxetine, or sertraline (considered together, +32%, P < 0.05), venlafaxine (+27%, P < 0.05), and inducers of P450 enzymes (−40%, P < 0.001). The final model included a tendency for higher concentration associated with female gender (+11%, P = 0.07) and accounted for 27% of observed interindividual variability. When considering a worst-case scenario, an elderly, nonsmoking woman prescribed fluvoxamine comedication was predicted to reach a 4.6-fold higher olanzapine concentration than a young male smoker coadministered carbamazepine. The current study suggests that patients characterized by a combination of factors associated with altered metabolism may benefit from olanzapine TDM.