Ground reaction force (GRF) measurements are often normalised to body mass to facilitate inter-individual comparisons. The objective of this study was to explore the effect of a rider on the GRFs and fetlock joint kinematics of trotting horses. The subjects were 5 dressage-trained horses and 3 experienced dressage riders. Ground reaction force measurements and sagittal view videotapes were recorded as the horses trotted at the same velocity in hand (3.49 +/- 0.52 m/s) and with a rider (3.49 +/- 0.46 m/s). Data were time-normalised to stance duration. Ground reaction force measurements were expressed in absolute terms and normalised to the system mass (horse or horse plus rider). All the horses showed changes in the same direction when comparing the ridden condition with the in-hand condition. There was an increase in the absolute peak vertical GRFs of the fore- and hindlimbs with a rider. However, the mass-normalised peak vertical GRFs were lower for the ridden condition, with the peak occurring later in the forelimbs and earlier in the hindlimbs compared with the inhand condition. Maximal fetlock angle and its time of occurrence were similar for the 2 conditions, but the fore fetlock joint was more extended during the later part of the stance phase in ridden horses. The presence of a rider appeared to affect the GRFs and fetlock joint kinematics differently in the fore- and hindlimbs, and the ridden horse did not seem to be equivalent to a proportionately larger horse. This should be considered when normalising for body mass in studies comparing horses in hand and ridden horses.