Background: Adults with mental illness may have difficulties with data collection methods such as questionnaires and accelerometry.

Aims: To assess the utility of questionnaires and accelerometry for assessing physical activity (PA) and sedentary behaviour (SB) in non-institutionalised adults with mental illness.

Methods: Participants were recruited from outpatient clinics and community organisations. Participants completed PA and SB questionnaires, wore accelerometers for 7 d, and rated the ease/difficulty of completing study components. Recruitment numbers, adherence, and ease/difficulty ratings were examined. Ease/difficulty ratings were compared between study components, and between participants by distress level.

Results: One hundred forty-two participants completed the questionnaires; they found it easier to report PA than reclining time (p = 0.017), and reclining time than sitting time (p < 0.001). Participants with high distress found it more difficult to report sitting time and PA than participants with low distress (p < 0.017). Ninety-nine participants (70%) completed the accelerometry; the majority (88%) met the minimum wear-time criteria. They found it easier to wear the monitor during the day than while sleeping (p < 0.001), and easier to complete accelerometry than questionnaires (p < 0.001).

Conclusions: Accelerometry was more feasible for assessing SB than questionnaires. Questionnaires were feasible for assessing PA, but less acceptable for people experiencing high distress.