In horses at a trot, the head moves up and down twice in one stride. In horses with unilateral forelimb lameness this movement is asymmetric. Computer-assisted kinematic analysis of vertical head movement can be used to quantify objectively lameness in horses in clinical trials. However, in mild lameness, absolute measurements of vertical head height may not be sensitive enough to detect small differences in lameness, and extraneous head movement by the horse due to curiosity, excitement or nervousness interferes with the accurate measurement of vertical head movement asymmetry. We describe a simple, signal-decompensation method of evaluating vertical head movement using a model of induced mild foot lameness in 9 horses. The technique assumes that the vertical head movement pattern can be broken down into 3 components; the vertical head movement caused by forelimb lameness (A1), the amplitude of the natural biphasic vertical head movement (A2) and extraneous head movement. Extraneous head movement is mathematically removed from the vertical head movement pattern. A1 and A2 are then calculated. After induction of lameness, mean A1 increased by 1.63 cm (range 0.10-3.33 cm, P = 0:005). Mean A2 did not significantly change after lameness induction. Error in reproduction of the original head movement pattern was 0.3-0.5%. We calculated that a hypothetical clinical trial would require 12 subjects for testing to be 80% certain that this difference would be successfully detected using this method of lameness evaluation.