Objectives: The purpose of this studywas to test a model of exercise behavior in older adults. It was hypothesized that gender, marital status, chronic illness, mental and physical health, self-efficacy and outcome expectations, fear of falling, and past exercise behavior were directly and/or indirectly associated with current exercise behavior. Methods:In this descriptive study, interviews were conducted with 201 older adults living in a continuing-care retirement community. Results:Twelve paths were significant, and the model accounted for 40% of the variance in exercise behavior. Self-efficacy expectations, outcome expectations, and prior exercise were directly associated with current exercise; health status, gender, and marital status were indirectly associated with current exercise behavior through self-efficacy and outcome expectations. Discussion:Recognizing and treating mental and physical health problems may directly influence self-efficacy and outcome expectations related to exercise. Moreover, interventions that strengthen self-efficacy and outcome expectations related to exercise may improve exercise behavior.