背景：運動對老年人維持或增進健康的影響已在許多文獻中有所討論，然而，只有三分之一的老年人從事規律運動。國內外許多研究探討運動行為影響因素，較少研究探討社區老年人運動持續性之相關因素。目的：瞭解社區老年人運動持續性之影響因素。方法：採橫斷性研究設計，以過去曾參與健走實驗研究的所有老年人為對象，以結構式問卷及面訪蒐集基本資料、健康狀況、運動自我效能、運動結果預期、知覺社會支持、規律運動行為、運動改變階段等資料進行統計分析。結果：共取得有效問卷108份，其中約76%的老年人執行規律運動，有規律運動並持續六個月或以上者佔68.5%。迴歸模式中以運動持續性為依變項時，被納入的自變項包括：慢性病罹患總數、運動自我效能、運動結果預期，整體解釋力為52.6%。結論/實務應用：本研究結果可提供未來擬定社區老年人運動介入計劃或執行衛生教育時之參考，藉由提昇老年人的運動自我效能、運動結果預期，並考量老年人慢性病罹患總數，或可有效提昇社區老年人規律運動持續性。

Background: The beneficial effects of exercise in older people have been well documented. However, only onethird of older people exercise on a regular basis. Most studies of exercise among older people have focused on the correlates of exercise behaviour. Research evidence related to factors that are associated with the maintenance of exercise in older adult populations is insufficient. Purpose: This study investigated the predictors of exercise maintenance behaviour among community‐dwelling older people. Methods: This cross‐sectional study used face‐to‐face interviews based on a set of structured questionnaires to collect data on participant demographic characteristics, health conditions, exercise self‐efficacy, exercise outcome expectancies, perceived social support, regular exercise behaviour, and exercise stage of change. All participants had taken part previously in an experimental health‐walk program and had given informed consent prior to participation in the present study. Results: A total of 76% of the participants engaged in regular exercise (n = 108), with 68.5% of this number engaging in regular exercise for over six months. Logistic regression analysis revealed several significant factors that influenced maintenance of exercise: total number of chronic diseases, exercise self‐efficacy, and exercise outcome expectancy, which collectively explained 52.6% of the total variance. Conclusions: The present study identified three factors (total number of chronic diseases, exercise self‐efficacy and exercise outcome expectancy) as crucial for explaining exercise maintenance in community older people. These factors should be considered when designing physical activity interventions and health education programs that are aimed to enhance the exercise maintenance behaviour in the community‐dwelling older adult population.