OBJECTIVE: This review evaluates the in vitro and in vivo evidence for inhibition of cytochrome P450 enzymes by the newer antidepressants and provides clinical recommendations for avoiding and managing drug interactions. METHOD: The international literature on the cytochrome P450 system and related drug interactions from 1966 to 1995 was reviewed. In vitro studies, pharmacokinetic trials in human subjects, and case reports were assessed. RESULTS: The newer antidepressants each inhibit a different cluster of cytochrome P450 enzymes, which are of relevance to the potential for drug interactions. Cytochrome P450 1A2 is inhibited by fluvoxamine and is implicated in drug interactions with theophylline, clozapine, and others. Fluoxetine, norfluoxetine, sertraline, and paroxetine are potent in vitro inhibitors of cytochrome P450 2D6 and are capable of causing marked elevations in plasma desipramine and nortriptyline concentrations. Fluoxetine, sertraline, and fluvoxamine are believed to inhibit cytochrome P450 2C because of observed interactions with phenytoin, diazepam, and other drugs metabolized by these enzymes. Cytochrome P450 3A4 metabolizes terfenadine, astemizole, carbamazepine, alprazolam, triazolam, and other benzodiazepines. Plasma concentrations of these drugs have increased when they are administered with fluvoxamine, nefazodone, fluoxetine, and sertraline. CONCLUSIONS: The majority of the newer antidepressants are associated with a risk for clinically significant drug interactions. A rapidly growing body of literature provides evidence for a distinct profile of cytochrome P450 inhibition and drug interaction risks by individual antidepressants. These findings underscore the need for definitive in vivo interaction studies of plasma from phenotyped patients treated with clinically effective antidepressant doses of medication, for direct comparative clinical studies, and for studies assessing the utility of phenotyping in clinical practice.