To find out whether botulinum toxin alters the excitability of cortical motor areas, we studied intracortical inhibition with transcranial magnetic stimulation in patients with upper limb dystonia before, 1 month after, and 3 months after the injection of botulinum toxin type A in the affected muscles. Eleven normal subjects and 12 patients with dystonia involving the upper limbs (7 with generalized dystonia, 2 with segmental dystonia, and 3 with focal dystonia) were studied. Patients were assessed clinically with the Dystonia Movement Scale. Paired magnetic stimuli were delivered by two Magstim 200 magnetic stimulators connected through a Bistim module to a figure-of-eight coil placed over the motor area of the forearm muscles. Paired stimulation was given at rest. A subthreshold (80% of motor threshold) conditioning stimulus was delivered 3 and 5 msec before the suprathreshold (120% of motor threshold) test stimulus. Electromyographic signals were recorded over the flexor or extensor muscles of the forearm on the affected side. We measured the amplitude of the test motor evoked potential (expressed as a percentage of the unconditioned motor evoked potential). All results were compared using ANOVA. In all patients, a botulinum toxin type A injection (50-100 mouse units) reduced dystonic movements in the arm. In normal subjects, electromyographic recordings showed significant inhibition of the test response. Before botulinum toxin injection, patients had less test response inhibition than normal subjects. One month after injection, patients had test response inhibition similar to that of normal subjects. At 3 months after injection, they again had less inhibition than normal subjects or patients at 1 month after injection. In conclusion, our data suggest that botulinum toxin can transiently alter the excitability of the cortical motor areas by reorganizing the inhibitory and excitatory intracortical circuits. The cortical changes probably originate through peripheral mechanisms.