OBJECTIVE: This study investigated the role of coping plans and the use of selection, optimisation and compensation (SOC) strategies within an experimental evaluation of a 26-week physical exercise intervention. METHODS: Older women (N=86, M (age)=73.7 years) were randomly assigned to a telephone-assisted or a self-administered coping planning intervention after 6 weeks' participation in an exercise programme. The number of different coping plans formulated, exercise-specific SOC strategy use and their interaction were used to predict objectively measured long-term adherence to the intervention. RESULTS: The number of coping plans formulated (beta=0.28) and goal-pursuit strategies reported (sum score of optimisation and compensation strategies, beta=0.39) predicted adherence to the exercise programme over 20 weeks. The predictive strength of coping plans increased with decreasing numbers of goal-pursuit strategies (beta=-0.21). Women supported via telephone reported significantly more coping plans than did women in the self-administered coping planning group, F(1,80)=9.47, p=0.003. CONCLUSION: Coping plans have a buffering effect on adherence levels when use of SOC strategies is low. Older women's adherence to physical activities may be improved if they are given direct support in generating coping plans involving strategies of selection, optimisation and compensation.