Objective: To determine whether there are differences in arm, leg and trunk activity measured by acceleration between weekdays and weekends in people undergoing rehabilitation in the subacute stage after stroke. Design: Cross-sectional study. Patients: Twenty-eight individuals with stroke (mean age 55.4 years; severe to mild impairment) and 10 healthy controls. Methods: A set of 5 3-axial accelerometers were used on the trunk, wrists and ankles during 2 48-h sessions at weekdays and over a weekend. Day-time acceleration raw data were expressed as the signal magnitude area. Asymmetry between the affected and less-affected limb was calculated as a ratio. Results: Participants with stroke used their both arms and legs less at weekends than on weekdays (p<0.05, effect size 0.32–0.57). Asymmetry between the affected and less-affected arm was greater at weekends (p < 0.05, effect size 0.32). All activity measures, apart from the less-affected arm on weekdays, were lower in stroke compared with controls (p<0.05, effect size 0.4–0.8). No statistically significant differences were detected between weekday and weekend activity for the control group. One-third of participants perceived the trunk sensor as inconvenient to wear. Conclusion: Increased focus needs to be applied on activities carried out during weekends at rehabilitation wards.