The linear estimation (LE) and spline surface Laplacian (SL) method were applied to single-trial EEG data. EEG was recorded in three subjects during voluntary, self-paced movements of the index finger. The EEG data were bandpass-filtered in the lower beta frequency range and showed short-lasting bursts of oscillations after termination of movement. These oscillations are termed postmovement beta synchronization. The realistic head geometry and the digitized positions of the electrodes were taken into account for accurate modeling of the anatomy. Regularization of the LE method was achieved by the truncated singular value decomposition. The LE and SL distribution of the postmovement beta synchronization showed similar spatial and temporal patterns. A clear increase of the LE source activity was found over the primary motor area. These results indicate that the postmovement beta synchronization is generated over the anterior bank of the central sulcus.