Two procedures for eliciting premovement potentials were compared: (1) the estimation of a 3 s interval elapsed after a warning auditory signal, and (2) classical "self-pacing". Eleven healthy right-handed subjects participated in the experiment, EEG records from scalp electrodes placed at Cz, C3+ and C4 + were analyzed. It has been shown that both procedures induced similar premovement potentials except that in the first procedure the early component of the potential was longer. The time estimation itself induced a negative slow potential consisting of a rapid set-up and a subsequent plateau.