Purpose: This study assessed the effectiveness of Active Start, a community-based behavior change and fitness program, designed to promote physical activity among sedentary community-dwelling older adults. Design and Methods: A quasi-experimental design was used. Data were analyzed using a within-group pretest-post-test design to calculate changes in the intervention group and changes in the waitlist comparison group. Further analyses were conducted after the comparison group completed the intervention. Paired "t" tests were employed to analyze unadjusted mean changes in physical performance measures from pretest to post-test. Repeated measures analysis of covariance (using SAS Proc Mixed) was then conducted to calculate the adjusted mean change for the intervention group. Results: Significant improvement was found on all performance measures (strength, flexibility, and balance) for the intervention group as a whole. Similar improvements were found among subgroups (Whites, African Americans, and Hispanics). No significant changes were found in the comparison group when they were in the control condition; however, they significantly improved on all measures after completing the intervention. Implications: This study suggests that a community-based physical activity program benefits sedentary, racially, and ethnically diverse older adults by coupling a behavioral change support group and fitness classes.