Grating onsets generate electrical brain responses with spatial distributions that indicate activation of the striate and extrastriate cortex for high and low spatial frequencies respectively (instantaneous electrical source analysis; Kenemans et al., 2000). To decrease overlap between electrical potentials originating from one or more ipsilateral and contralateral sources we studied the effect of hemifield stimulus presentation, next to full-field. Checkerboard stimuli with spatial frequencies of 0.53 (low), 1.06 (middle) and 4.25 (high) cycles per degree were presented to 10 participants. Spatiotemporal source analysis of checkerboard evoked responses indicated that regardless of spatial frequency first the contralateral extrastriate cortex was activated, next the contralateral striate and finally the ipsilateral striate cortex. With full-field stimulation bilateral extrastriate activity was followed by bilateral striate activity. Furthermore, we extended the latency-of-best-fit test procedure (Kenemans et al., 2002) from instantaneous to spatiotemporal dipole models. This procedure involves the estimation of individual spatial and temporal source parameters, starting from the grand average dipole solution. In a next step these parameters are statistically analyzed by (M)ANOVA. Preliminary results of such analyses confirm the above conclusions, that were based on visual inspection. We will elaborate these by a detailed statistical analysis of both spatial and temporal source parameters of full-field and hemifield visual evoked potentials for checkerboards of low, intermediate and high spatial frequency. We will inspect these for an explanation of the apparent discrepancies between these and earlier dipole solutions for grating onsets. (Supported by Netherlands Organization for Scientific Research, NWO, grant 575-25-015).