BACKGROUND/OBJECTIVES: Fatness and fitness are associated with physical activity (PA) but less is known about the prospective associations of adiposity and muscle strength with PA. This study aimed to determine longitudinal associations of body mass index (BMI), waist circumference (WC) and grip strength (GS) with objectively measured PA. SUBJECTS/METHODS: Data are from the UK Biobank study. At baseline (2006-2010), BMI, WC and GS were objectively measured. At follow-up (2013-2015), a sub-sample of 93 015 participants (52 161 women) wore a tri-axial accelerometer on the dominant wrist for 7 days. Linear regression was performed to investigate longitudinal associations of standardised BMI, WC and GS at baseline with moderate-to-vigorous PA (MVPA) and acceleration after a median 5.7-years follow-up (interquartile range: 4.9-6.5 years). RESULTS: Linear regression revealed strong inverse associations for BMI and WC, and positive associations for GS with follow-up PA; in women, MVPA ranges from lowest to highest quintiles of GS were 42-48 min day(-1) in severely obese (BMI⩾35 kg m(-)(2)), 52-57 min day(-1) in obese (30⩽BMI<35 kg m(-)(2)), 61-65 min day(-1) in overweight (25⩽BMI<30 kg m(-)(2)) and 69-75 min day(-1) in normal weight (18.5⩽BMI<25 kg m(-2)). Follow-up MVPA was also lower in the lowest GS quintile (42-69 min day(-1)) compared with the highest GS quintile (48-75 min day(-1)) across BMI categories in women. The pattern of these associations was generally consistent for men, and in analyses using WC and mean acceleration as exposure and outcome, respectively. CONCLUSIONS: More pronounced obesity and poor strength at baseline independently predict lower activity levels at follow-up. Interventions and policies should aim to improve body composition and muscle strength to promote active living.International Journal of Obesity advance online publication, 6 June 2017; doi:10.1038/ijo.2017.122.