BACKGROUND Cervical spinal cord dysfunction is a common problem in equine medicine and the currently available tests give no objective information about the functionality of the nervous tracts. Therefore, transcranial magnetic stimulation (TMS) was performed in 84 healthy horses of different height in order to have an objective measure for the integrity of the descending motor tracts in normal horses. OBJECTIVE To obtain reference values for onset latency and peak-to-peak amplitude of magnetic motor evoked potentials (MMEPs) and to evaluate the possible effect of height, age and gender on the neurophysiological measures. METHODS All horses were sedated and stimulated transcranially by using a magnetic coil placed on the forehead. The stimulator triggered the sweep of an electromyogram machine that recorded MMEPs bilaterally from needle electrodes in the extensor carpi radialis and cranial tibial muscles. In that way, it was possible to measure latency between stimulus and onset of response. RESULTS A significant difference was found between recordings made in the fore- and hindlimbs; MMEPs recorded in the front legs had a shorter onset latency and higher peak-to-peak amplitude. Mean +/- s.d. normal values for onset latency of 19.32 +/- 2.50 and 30.54 +/- 5.28 msecs and peak-to-peak amplitude values of 9.54 +/- 3.73 and 6.62 +/- 3.62 mV were obtained for extensor carpi radialis and cranial tibial muscles, respectively. The left-to-right difference in onset latency and peak-to-peak amplitude was not significant. In the same horse, differences up to 0.82 and 1.53 msecs for the extensor carpi radialis and cranial tibial muscles, respectively, lie within the 95% confidence limit and are considered normal. In contrast to onset latency, peak-to-peak amplitude showed a very large intra- and interindividual variability, even in the same muscle. To reduce the variability and predict normal values of new individual cases, influence of height, weight, age and sex on the MMEPs were determined. No significant effects of sex were observed on onset latency and peak-to-peak amplitude. The age of the horse had only a small but significant effect on peak-to-peak amplitude, with larger responses in older horses. Height at the withers and weight of the horse, parameters that strongly correlate with the size of the horse, had an important significant influence on onset latency but not on peak-to-peak amplitude. The age of the horse and height at the withers were used to predict peak-to-peak amplitude and onset latency, respectively, in normal horses. CONCLUSIONS TMS is an excellent addition to the few tools we have for noninvasive imaging of the equine nervous system. Magnetically evoked potentials are highly reproducible and recent advances suggest that the applications of TMS in horses will continue to grow rapidly.