OBJECTIVE: To observe patterns of antidepressant use and concomitant psychotropics to determine implications for clinical prescribing practice. METHOD: Among 1.6 million members of Merck-Medco Managed Care, L.L.C. followed continuously between 1/1/96 and 12/31/98, we identified all patients (N=42,510) who received a new prescription (defined as none within the prior 12 months) for an antidepressant in 1997 (Index AD). We observed the time between the use of other concomitant psychotropics (e.g. antipsychotics(AP)/atypicals(AP-A), anxiolytics (ANX)/buspirone(ANX-B), sedative/hypnotics (SH) and zolipidem (SH-Z), and miscellaneous (MISC)) in the year prior to or after the index AD prescription. RESULTS: 14, 792 (34.8%) of AD patients used a concomitant psychotropic. These patients displayed a parallel pattern in use of all classes of concomitants: 0.2-12.2% using at least one additional class 4 months prior to, 0.3-2.0% same day, and 0.8-3.0% 4 months after the index antidepressant. However, patients were more likely to receive a traditional antipsychotic and anxiolytic prior to the index AD, and an atypical antipsychotic and buspirone after the index AD. Concomitant % of AD patients Pre Index AD Same Day Post Index AD AP 2.3 .7 .3 1.3 AP-A 2.4 .2 .4 1.8 ANX 21.3 12.2 2.0 7.1 ANX-B 3.7 1.5 .4 1.8 MISC 2.6 .7 .3 .8 SH 5.7 3.0 .4 2.3 SH-Z 9.0 4.4 .7 3.0 CONCLUSION: Whereas we hypothesized excessive benzodiazepine and hypnotic usage pre and post index AD, we noted lower than expected usage. We did not expect the pattern of increased atypical antipsychotic usage on the same day or subsequent to an index AD. All classes of concomitant psychotropics revealed a similar general pattern: a steady increase in the 4 months prior to an index AD, a peak on the index AD date, and a tapering over the next 4 months. We speculate that continuing medical education of physicians may be contributing to a decrease in the misuse of benzodiazepines and sedative/hypnotics.