The benefits of regular exercise for older adults are well documented and include improvements in physical, functional, as well as psychological, health. The purpose of this descriptive study was to test a theoretically and empirically based model describing the factors that influence exercise behaviour of older adults in the United States of America. The hypothesized model suggested that age, gender, and mental and physical health have an effect on self-efficacy and outcome expectations, and that all these variables influence exercise behaviour. Exercise behaviour was hypothesized to have a reciprocal relationship with self-efficacy expectations and mental and physical health. The convenience sample was 187 older adults living in a continuing care retirement community (CCRC) in Baltimore, Maryland. A one-time health interview was conducted which included a measure of self-efficacy and outcome expectations related to exercise, a measure of health status (SF-12), and gathering of information from participants about their actual exercise behaviour. Of the 187, 71 (38%) reported participating in 20 minutes of continuous aerobic exercise at least three times per week over the previous 3 months. Six hypothesized paths were significant. The model fitted the data and accounted for 32% of the variance in exercise behaviour. Interventions that focus on strengthening self-efficacy and outcome expectations can improve exercise behaviour in older adults.