___________________________________________________________________The life expectancy of adults with mental illness is significantly less than that of the general population. This is largely due to poor physical health. Physical activity is consistently recommended for the prevention and management of noncommunicable diseases and also has mental health benefits. The aim of this thesis was to understand and promote physical activity in adults with mental illness, to improve physical health.Study One was a cross-sectional study of inpatients in a private hospital. It assessed the (i) feasibility of self-report and objective measurement of physical activity and sedentary behaviour, (ii) levels of physical activity and sedentary behaviour, and (iii) physical activity attitudes and preferences for contexts and sources of support. 101 participants completed questionnaires on physical activity and sitting time, activity preferences and attitudes, psychological distress and sociodemographic and health variables. 38 also wore an accelerometer for 7 consecutive days.Feasibility of measurement was assessed in terms of participant engagement; selfreported ease/difficulty; extreme self-report data values; and adherence to accelerometer wear time criteria. Findings demonstrated that inpatient adults with mental illness can engage with both questionnaire and accelerometry measurement, that it was more feasible but less acceptable to wear an accelerometer than to complete questionnaires, and that this was not influenced by level of psychological distress.Questionnaire data were used to determine time spent in (i) walking and moderateand vigorous-intensity activity (MVPA), and (ii) domain specific sitting time.Accelerometry was used to determine mean daily time spent in MVPA and sedentary behaviour. Bivariate associations between self-reported MVPA, sedentary behaviour and explanatory variables of gender, age, education, body mass and distress were analysed using regression analyses. Self-report data indicated a median of 32 minutes/day in MVPA and a median of 761 minutes/day in sedentary behaviour.iii Accelerometry data indicated an average of 37 minutes/day in MVPA and 664 minutes/day in sedentary behaviour. Analyses indicated no significant associations between explanatory variables and MVPA or sedentary behaviour.Questionnaire data were used to determine (i) physical activity interest; (ii) reasons to do activity; (iii) general knowledge regarding activity benefits; (iv) preferences for activity type, context and sources of support; and (v) activity barriers. More than three quarters (77%) of participants expressed high interest to do physical activity while in hospital, with the most common reasons being to maintain physical health and improve emotional wellbeing (≥95%). More than 90% of participants agreed physical activity was beneficial for managing psychological wellbeing, heart disease, stress, diabetes and quality of life; but fewer than half agreed that activity had benefits for serious mental illness. Part...