The pattern electroretinogram was recorded to checkerboard stimuli with a wide range of check sizes and two stimulus field sizes. Check sizes ranged from 0.25 degree to 7 degrees (field size, 16 degrees x 14 degrees) and 0.25 degree to 15 degrees (field size, 32 degrees x 27 degrees) in 14 and seven subjects, respectively. Reversal rate was 4.5/s. For minimal intrusion of blink artifacts the interrupted stimulation technique was employed. The P50 and N95 components of the pattern electroretinogram were evaluated separately. With both stimulus field sizes amplitude of P50 and N95 was maximal between 0.75 degree and 1 degree. With smaller check sizes the amplitude dropped monotonically. With larger check sizes field size played a role: with the 16 degrees x 14 degrees field, P50 gradually dropped to 89% from 1 degree to 7 degrees, which was paralleled by N95 only up to 7 degrees, where N95 dropped to 81% (p < 0.05). With the 32 degrees x 27 degrees field, there was no significant difference in size dependency between P50 and N95 for large check, both components staying constant from 1 degree to 15 degrees. We conclude that there is only minor large-check attenuation of the pattern electroretinogram, especially with a large field. The apparent field-size dependency may explain previous discrepancies in the literature.