Housing conditions for horses impose various levels of confinement, which may compromise welfare. Lying behaviour and activity can be used as welfare indicators for domestic animals and rebound behaviour suggests a build-up of motivation resulting from deprivation. The objective of this study was to determine if activity and lying behaviour of horses are affected by housing conditions and to investigate the occurrence of rebound behaviour after release from confinement. Eight horses were subjected, in pairs, to each of four experimental treatments; paddock (P), fully stabled (FS), partly stabled (PS) and yard (Y). Each horse received 6 days acclimatisation prior to the 24 h recording period. Time spent in lying and activity were electronically recorded using a tilt switch and motion sensor connected to a data logger worn on the horse's left foreleg. Time spent active during the first 5 min of release from stable to paddock in the PS treatment (days 1 and 5) and at the same time of day in the P treatment was used as a measure of rebound behaviour. Effect of housing conditions on total time spent active was highly significant (FS 5 123 s, PS 5 158 s, Y 5 377 s, P 5 779 s, P , 0.001). Housing conditions did not significantly affect total time spent lying (P 5 0.646). Horses were significantly more active, compared with baseline paddock behaviour, on release from stabling on both days 1 (P 5 0.006) and 5 (P 5 0.025) of PS treatment. These results suggest that activity patterns of horses, but not lying behaviour, are affected by the housing conditions tested and that rebound activity occurs in horses after a period of confinement.