We used transcranial magnetic stimulation to study the modulation of motor cortex excitability after rapid repetitive movements. Eleven healthy subjects aged 24-32 years were evaluated. Serial motor-evoked potential (MEP) recordings were performed from the right thenar eminence every 5 min for a period of 20 min at rest and for a period of 35 min after repetitive abduction-adduction of the thumb at maximal frequency for 1 min. All subjects presented distinct changes in MEP amplitude after exercise with an approximately 55% mean maximal decrease compared with basal conditions and complete recovery 35 min after the end of the exercise. The time course of MEP amplitude changes presented the following trend: (1) a rapid decrease phase within the first 5 min; (2) a maximal depression phase of 10 min duration (from the 5th to the 15th min); and (3) a slow recovery phase. No significant modifications in post-exercise MEP amplitude were found in ipsilateral non-exercised muscles. In order to determine the level where these changes take place, we recorded the M and F waves induced by median nerve stimulation at the wrist (all subjects) and MEPs in response to transcranial electrical stimulation (five subjects) at rest and during the decrease and maximal depression phases. None of these tests were significantly affected by exercise, indicating that the motor cortex was the site of change. Evaluation of maps of cortical outputs to the target muscle, performed in four subjects, showed an approximately 40% spatial reduction in stimulation sites evoking a motor response during the maximal depression phase. These data prove that exercise induces a reversible, long-standing depression of cortical excitability, probably related to intracortical presynaptic modulation, which transitorily reduces the motor representation area.