OBJECTIVES: The authors analyzed self-reported questionnaire data from the 1987 National Medical Expenditure Survey (NMES) to determine the smoking patterns of veterans. METHODS: Using NMES data, the authors compared veterans versus nonveterans overall, women veterans versus women nonveterans, Vietnam-era veterans versus other veterans, and veterans whose usual source of medical care was the Department of Veterans Affairs (VA) system versus veterans who received care elsewhere. RESULTS: The likelihood of ever having smoked cigarettes was higher for veterans than for nonveterans and for women veterans than for women nonveterans. The prevalence of current smoking was higher for veterans than for nonveterans and higher for those seeking care within the VA system than for other veterans. CONCLUSIONS: Given the enormous health care costs associated with smoking, health promotion efforts should be developed to reduce the high rate of smoking among veterans--especially those who are consumers of VA health care.