The present study aimed at investigating whether different aspects of performance on a prospective reproduction task with a standard duration of 2 s had different antecedents in the event-related potentials (ERPs) accompanying standard presentations and reproduced intervals. On each trial, first the standard duration was presented as an empty auditory interval. After a short delay, participants reproduced this interval by means of two button presses defining onset and offset of the reproduced interval. About 25 participants were divided into groups of poor and good time estimators, once based on the coefficient of variation of their reproduced durations, and once based on the absolute error score. Interestingly, for both performance measures differences between participants were accompanied by ERP differences during standard presentations rather than reproduction itself. While larger P300 amplitudes evoked by standard interval offset predicted good reproduction performance in terms of a small coefficient of variation, small absolute error scores were accompanied by larger fronto-central negative slow wave during standard presentations. The results suggest an important role of attention to both the critical events that delimit the to-be-estimated temporal intervals, and to the passing of time between these events.
