A horse's legs are compressed during the stance phase, storing and then returning elastic strain energy in springlike muscle-tendon units. The arrangement of the muscle-tendon units around the lever-like joints means that as the leg shortens the muscle-tendon units are stretched. The forelimb anatomy means that the leg can be conceptually divided into two springs: the proximal spring, from the scapula to the elbow, and the distal spring, from the elbow to the foot. In this paper we report the results of a series of experiments testing the hypothesis that there is minimal scope for muscle contraction in either spring to adjust limb compliance. Firstly, we demonstrate that the distal, passive leg spring changes length by 127 mm (range 106-128 mm) at gallop and the proximal spring by 12 mm (9-15 mm). Secondly, we demonstrate that there is a linear relationship between limb force and metacarpophalangeal (MCP) joint angle that is minimally influenced by digital flexor muscle activation in vitro or as a function of gait in vivo. Finally, we determined the relationship between MCP joint angle and vertical ground-reaction force at trot and then predicted the forelimb peak vertical ground-reaction force during a 12 m s(-1) gallop on a treadmill. These were 12.79 N kg(-1) body mass (BM) (range 12.07-13.73 N kg(-1) BM) for the lead forelimb and 15.23 N kg(-1) BM(13.51-17.10 N kg(-1) BM) for the non-lead forelimb.