Transcranial magnetic stimulation (TMS) of the motor cortex of one hemisphere (conditioning stimulus (CS)) inhibits EMG responses evoked in distal hand muscles by a later magnetic stimulus given at an appropriate interval, over the opposite hemisphere (test stimulus (TS)). This effect is commonly attributed to an inhibition produced at cortical level via a transcallosal route. The present study assessed the reproducibility of the transcallosal inhibition effects in different sessions in healthy subjects. Within- and between-subject variability, relating to interhemispheric differences was also evaluated. A magnetic CS on one hemisphere effectively inhibited EMG responses of the abductor digiti minimi stimulated by a TS delivered over the opposite hemisphere in a range of intervals centered at 12 ms. Even though group effects were reproduced in separate sessions, the high between- and within-subject variability yielded low test-retest correlations. This differentiation forces the definition of reproducibility (or repeatability), as the replication of the same mean curves of EMG reduction, and of reliability, as the between- or within-subject correlations between values of specific EMG measures. (C) 2003 Elsevier Science Ireland Ltd and the Japan Neuroscience Society. All rights reserved.