Abstract Transcranial magnetic stimulation (TMS) results in a series of deflections in electroencephalography (EEG) recordings known as a TMS-evoked potential (TEP). However, it remains unclear whether these responses reflect neural activity resulting from transcranial stimulation of the cortex, the sensory experiences of TMS, or a combination of the two. Across three experiments (total n = 135), we recorded EEG activity following TMS to the dorsolateral prefrontal cortex, premotor cortex, and parietal cortex as well as a sensory control condition (stimulation of the shoulder or electrical stimulation of the scalp with a click sound). We found that TEPs showed a stereotypical frontocentral N100/P200 complex following TMS of all cortical sites and control conditions, regardless of TMS intensity or the type of sensory control. In contrast, earlier TEPs (<60 ms) showed site-specific characteristics which were largest at the site of stimulation, although TEP topographies were distorted in a subgroup of individuals due to residual TMS-evoked muscle artifact despite cleaning with independent component analysis. Self-reported sensory experiences differed across sites, with prefrontal stimulation resulting in stronger auditory (click sound perception) and somatosensory input (scalp muscle twitch, discomfort) than premotor or parietal stimulation, a pattern that was reflected in the amplitude of later (N100/P200), but not earlier (<60 ms) TEP peak amplitudes. Later TEPs were also larger in individuals who experienced stronger click sound perception and, to a lesser extent, TMS-evoked scalp muscle twitches. Increasing click sound perception by removing auditory masking increased N100/P200 amplitudes without altering earlier peaks, an effect which was more prominent at sites with more successful masking. Together, these findings suggest that the frontocentral N100/P200 complex primarily represents a generalised sensory response resulting from TMS-related auditory and somatosensory input when present. In contrast, early TEP peaks likely primarily reflect activity resulting from transcranial stimulation of the cortex when artifacts were adequately accounted for. The results have important implications for designing and interpreting TEP studies, especially when comparing TEPs between stimulation sites and participant groups showing differences in sensory experiences following TMS.