OBJECTIVE: The authors pilot a smoking-cessation outreach for the homeless that extends medical students' tobacco cessation education. METHOD: In this prospective study, second-year medical students administered cognitive behavior therapy or unstructured support to homeless subjects to help them quit smoking. Self-report and biological measures (carbon monoxide) of smoking taken at baseline and follow-up were analyzed using t tests to determine intervention efficacy. RESULTS: Out of 11 enrolled subjects, six completed the protocol and all decreased their smoking frequency. The mean rate of smoking dropped significantly from 19 to nine cigarettes per day when pooling all subjects, and carbon monoxide mean level decreased from 28.0 to 20.2. CONCLUSIONS: The homeless subjects who received counseling from medical students significantly reduced their smoking frequency. Subject recruitment and retention were challenges, but a close partnership with local homeless shelters and the addition of pharmacotherapy could improve outcomes and are recommended for future efforts.