Hemispheric processing differences were assessed by presenting square matrices that varied in size and the number of filled-in cells. Subjects judged whether the matrix contained an even or odd number of filled cells. Experiment 1 employed relatively small matrix sizes (2 x 2, 3 x 3, and 4 x 4), and Experiment 2 employed relatively large matrix sizes (4 x 4, 6 x 6, and 8 x 8). Response time was shorter and error rates lower for left visual field/right hemisphere (LVF/RH) presentations compared to right visual field/left hemisphere (RVF/LH) presentations, with the larger matrices demonstrating the strongest visual field/hemispheric effects. Increases in the number of filled cells contributed to increases for the LVF/RH response time advantage only for the larger arrays. Analysis of the data from both studies collapsed across the number of filled cells produced highly consistent LVF/RH advantages for both response time and error rate, with stronger LVF/RH advantages found for the larger matrix sizes of both studies. The findings suggest that visual stimulus spatial frequency is a key determinant of hemispheric processing advantages, but that this factor is constrained by stimulus size variation. Theoretical implications with respect to the hemispheric processing double filtering by frequency model are discussed.