For 35 years, some researchers have argued that CNV resolution may affect or even produce the increased P3 for NoGo compared to Go trials, and thus that no ‘inhibitory’ NoGo P3 exists. This is based on the work of Simson et al. (1977b), the scalp topography of potentials in auditory and visual Go/NoGo tasks. Electroencephalography and Clinical Neurophysiology, 43, 864–875, which compared Go and NoGo topography after CNV was subtracted from NoGo trials only. Specifically, the NoGo P3 topography showed the distinctive frontocentral maximum, which is often linked to motor inhibition, when referenced to a pre-target baseline. This NoGo topography changed to a more parietal maximum, similar to that on Go trials, when referenced to a pre-cue baseline. Many researchers have cited this study, while failing to use the delayed response design on which Simson et al. based their argument. We attempted to replicate Simson et al.'s experiment with delayed responses and also with immediate responses, as are more often used. As expected, the amplitudes of CNV and P3 to both Go and NoGo trials were increased when immediate compared to delayed responses were required, but we failed to replicate the topographic shift of NoGo P3 with different baselines for both delayed and immediate responses. That is, subtraction of the CNV from NoGo P3 did not change the distinctive frontocentral topography of this component. The results suggest that CNV may affect the amplitude and measurement of the NoGo P3, but that NoGo P3 anteriorisation is not caused by CNV resolution.