Background: Evidence suggests that physicians are less likely to identify alcohol problems in females than in males.

Purpose: To compare the performance of family medicine residents with male and female simulated patients (SPs) posing as problem drinkers.

Methods: Fifty-six family medicine residents completed a baseline survey on knowledge and attitudes towards problem drinkers. Each resident was then visited by one male and female unannounced SP. The male and female roles were similar with respect to presenting complaint (insomnia or hypertension), age, social class, and drinking history.

Results: Residents expressed slightly more positive attitudes towards female than male patients (3.32 vs. 3.09, p < .001). Residents scored higher with undetected male than with undetected female SPs on the assessment checklist (5.1 vs. 3.2, p < .045), the management checklist (4.4 vs. 3.2, p = .032), and an interpersonal rating scale (the Alcohol Skills Rating Form; 5.5 vs. 4.7, p = .023).

Conclusion: Educational programs should focus on improving physicians' clinical skills in the identification and treatment of alcohol problems in women.