Repetitive transcranial magnetic stimulation (rTMS) holds promise as a probe into the pathophysiology and possible treatment of neuropsychiatric disorders. To explore its regional effects, we combined rTMS with positron emission tomography (PET). Fourteen healthy volunteers participated in a baseline 18-fluorodeoxyglucose (FDG) PET scan. During a second FDG infusion on the same day, seven subjects received 30 min of 1 Hz rTMS at 80% of motor threshold to left prefrontal cortex, and seven other subjects received sham rTMS under identical conditions. Global and normalized regional cerebral glucose metabolic rates (rCMRglu) from the active and sham conditions were compared to baseline and then to each other. Sham, but not active 1 Hz rTMS, was associated with significantly increased global CMRglu. Compared to baseline, active rTMS induced normalized decreases in rCMRglu in right prefrontal cortex, bilateral anterior cingulate, basal ganglia (L>R), hypothalamus, midbrain, and cerebellum. Increases in rCMRglu were seen in bilateral posterior temporal and occipital cortices. Sham rTMS compared to baseline resulted in isolated normalized decreases in rCMRglu in left dorsal anterior cingulate and left basal ganglia, and increases in posterior association and occiptal regions. Differences between the 1 Hz active versus sham changes from baseline revealed that active rTMS induced relative decrements in rCMRglu in the left superior frontal gyrus and increases in the cuneus (L>R). One Hertz rTMS at 80% motor threshold over the left prefrontal cortex in healthy subjects compared to sham rTMS in another group (each compared to baseline) induced an area of decreased normalized left prefrontal rCMRglu not directly under the stimulation site, as well as increases in occipital cortex. While these results are in the predicted direction, further studies using other designs and higher intensities and frequencies of rTMS are indicated to better describe the local and distant changes induced by rTMS.