REASONS FOR PERFORMING STUDY: Inadequate track surfaces are believed to be a risk factor in the occurrence of musculoskeletal injuries, but quantification of the shocks and vibrations provoked by hoof impact on different ground surfaces (including new synthetic tracks) has been insufficiently documented in trotters under high-speed training conditions. OBJECTIVES: To test the reliability and sensitivity of an accelerometric device to discriminate between the biomechanical effects of 2 different tracks at high speed. METHODS: Three French Trotters were used and their right front hooves were equipped with one triaxial accelerometer. Two different track surfaces (crushed sand track: S and all-weather waxed track: W) were tested when horses were trotting in a straight line. For each session of measurements, trials were repeated 3 times in a Latin square design. The speed of the runs was set at 10 m/s, controlled by the driver and recorded synchronously. Sample rate was set at 6 kHz. Acceleration of the hoof (resultant vector and 3D components), power spectral density at impact and variability (between strides, trials, sessions and horses) were analysed. Statistical differences were tested using a GLM procedure (SAS). Least square mean differences were used for comparisons between tracks (P < 0.05). RESULTS: Results showed that the deceleration of the hoof (magnitude of the resultant vector) was statistically different between the 2 tracks with an attenuation of the shock of about 50% on the all-weather waxed track. Magnitude of the power spectral density was reduced at higher frequencies on W. CONCLUSIONS AND CLINICAL RELEVANCE: These preliminary results demonstrate the sensitivity of the tool to discriminate between the different behaviours of the hoof on the different track surfaces at high speed. Deceleration and vibration of the hoof at impact were reduced on W compared to S, suggesting a better shock-absorbing quality of this track.