Older, sedentary, urban-living, ethnic minority women are at high risk for preventable disease, but it is difficult to engage this population in health promotion efforts. This study tested two methods of engaging Hispanic and African American women, who were at high risk for cardiovascular disease, in a 10-week aerobic fitness program. The program was offered to 76 participants, in either a women's health clinic or a church. Attendance was the primary dependent variable and was recorded at each exercise session. Other variables, including the Baecke Questionnaire of Habitual Physical Activity, Fat Frequency Questionnaire, Self-Efficacy for Exercise Behaviors Scale, Social Support and Exercise Survey, and Psychological General Well-Being Schedule, were measured prior to the intervention, at the end of the 10-week program, and at 3-month follow-up. Age predicted attendance, independently of site. Women in the highest age quartile (50 - 70 years) attended more than twice as many exercise sessions compared to women in the lowest age quartile (17 - 27 years). The relationship between older age and attendance was particularly strong for Hispanic women. Church parishioners were primarily women over the age of 40, making it impossible to disentangle the relative effect of locale. These findings are relevant for clinicians who design exercise programs targeting older, ethnic, minority women. Administrators who design exercise programs for urban-living women should consider age of the target population when selecting the most conducive setting.