Background. If a simple system of instrumented monitoring was possible early after stroke, therapists may be able to more readily gather information about activity and monitor progress over time. Our aim was to establish whether a device containing a dual-axis accelerometer provides similar information to behavioural mapping on physical activity patterns early after stroke. Methods. Twenty participants with recent stroke ≤2 weeks and aged >18 were recruited and monitored at an acute stroke ward. The monitoring device (attached to the unaffected leg) and behavioural mapping (observation) were simultaneously applied from 8 a.m. to 5 p.m. Both methods recorded the time participants spent lying, sitting, and upright. Results. The median percentage and interquartile range (IQR) of time spent lying, sitting, and upright recorded by the device were 36% (15–68), 51% (28–72), and 2% (1–5), respectively. Agreement between the methods was substantial: Intraclass Correlation Coefficient (95% CI): lying 0.74 (0.46–0.89), sitting 0.68 (0.36–0.86), and upright 0.72 (0.43–0.88). Conclusion. Patients are inactive in an acute stroke setting. In acute stroke, estimates of time spent lying, sitting, and upright measured by a device are valid.