This study used repetitive transcranial magnetic stimulation (rTMS) to investigate the roles of the right dorsolateral prefrontal cortex (DLPFC) and supplementary motor area (SMA) in short (500 ms) and long (2 s) interval timing. The results were compared with rTMS over the leg area of motor cortex, an area not thought to be involved with time estimation. rTMS was delivered during one of two phases of a time reproduction task: at the onset of the Estimation Phase (presentation of the interval to be timed) and at the onset of the Reproduction Phase (subjects' reproduction of the timed interval). There was a significant main effect of Site (SMA vs. right DLPFC vs. leg motor area) due to the fact that rTMS over the right DLPFC caused subjects to underestimate time intervals compared with rTMS over the leg motor area. There was also a significant three-way interaction between Site, Duration and Phase (Estimation Phase vs. Reproduction Phase) that post hoc analyses showed was due to underestimation of long intervals when rTMS was given over the right DLPFC at the start of the Reproduction Phase. There was no effect of rTMS over the right DLPFC or SMA in the short interval task. This is consistent with previous studies showing that the right DLPFC is important in estimating time intervals in the seconds-range. In addition, we suggest that the selectivity of the rTMS effect for the Reproduction Phase indicates that the right DLPFC plays a particular role in memory processes.