In motor learning, people must process movement errors when either adapting well-known movements or acquiring new motor skills (de novo learning). While these two motor learning types are distinct behaviorally, we understand less about their underlying neural mechanisms. Here, we investigated event-related potentials (ERPs) during movement preparation and outcome in reaching movements. We distinguished the two motor learning types by having participants (N = 32) train with two perturbations: a 30° visuomotor rotation (adaptation) and a mirror reversal (de novo learning). Participants learned to compensate for both perturbation types. For movement outcome, we time-locked to feedback onset at the end of the movement. We found a negative-going ERP before feedback onset, with perturbed reaches showing more negativity than aligned baseline reaches. However, the ERP amplitude did not scale with perturbation type nor error magnitude. Furthermore, we found a larger P3 component, or more attention allocation, after feedback onset in the perturbed conditions. For movement preparation, we time-locked to the go signal onset before the movement. We found a Readiness Potential that was dependent on the participant’s intended movement direction. We then quantified changes in movement preparation per perturbation type across learning, using a Lateralized Readiness Potential. However, we found no LRP differences between the fixed rotation and mirror reversed reaches. Thus, although the ERPs we investigated represent movement preparation and outcome error processing, these markers are unable to distinguish between motor adaptation and de novo learning. Other markers in relation to these two motor learning types warrant further investigation.