One third of adults over 65 yr old fall each year. Wide-ranging consequences include fracture, reduced activity, and death. Research synthesis suggests that falls-prevention programs can be effective in reducing falls by about 20%. Strength and balance training is the most efficacious component, and the assumed method of effect is an improvement in these performance domains. There is some evidence for this, but the authors have previously proposed an alternative method, activity restriction, leading to a reduction in subsequent falls through a reduction in exposure. The aim of this study was to examine physical activity in older fallers, applying a theory of adaptation, to ascertain predictors of habitual physical activity. Referrals to hospital- and community-based exercise programs were assessed for (a) habitual walking steps and (b) coping strategies, falls self-efficacy, social support, and balance mobility. There was no average group change in physical activity. There was high interindividual variability. Two coping strategies, loss-based selection and optimization, best explained the change in physical activity between baseline and follow-up. Notwithstanding some limitations, this work suggests further use of adaptation theory in falls research. A potential application is the creation of a profiling tool to enable clinicians to better match treatment to patient.