Five sound mature horses, age 8-14 years, with toe angles 45.5-55.0 degrees were placed on a specially designed platform with only the left forelimb weightbearing, which allowed the successive production of 7 different conditions of foot orientation: neutral position (N), 5, 10, 15 degrees heel (H5, H10, H15) and toe (T5, T10, T15) elevation, performed according to 2 different sequences. For each condition, 2 lateromedial radiographs were taken to evaluate the metacarpophalangeal joint (MPJ) and both interphalangeal joint (PIPJ and DIPJ) angles, respectively. In addition, two-dimensional (2-D) kinematic recordings, using reflective skin markers placed laterally on the left forelimb joints, were performed. The value of each joint angle (JA) and its angular variation to N (AV) were considered for statistical analysis. For all JA and their AV determined radiographically, the condition effect was significant and the sequence did not influence the differences between conditions. The relationships between the JA and the conditions (in the sequence: T15, T10, T5, N, H5, H10, H15) were overall linear for the 3 joints. The maximal range T15-H15 was mean +/- s.d. 6.9 +/- 2.0 degrees for MPJ (dorsal angle decrease), 7.3 +/- 1.0 degrees and 29.5 +/- 1.8 degrees for PIPJ and DIPJ (palmar angle decrease), respectively. Contrary to JA and in spite of large variations in hoof conformation, AV did not show any significant horse effect. Wide differences were observed between the digital JA measured radiographically and from kinematics, which could be related to cutaneous displacement. Heel elevation induced elbow flexion, whereas a slight extension was observed with toe elevation. These movements, even slight (mean elbow angle amplitude: 3.6 +/- 1.3 degrees), may have consequences upon flexor tendon tension.