Pattern electroretinograms were recorded to phase-reversing checkerboard stimuli with DTL electrodes under conditions close to those of the ISCEV pattern electroretinogram guidelines. Both transient (2 reversals/s) and steady-state (16 reversals/s) stimulation was used. The check sizes were 0.4 degree, 0.8 degree and 16 degrees; the mean luminance 45 cd/m2, the contrast 98%, and the field size 32 degrees x 27 degrees. In 42 eyes of 21 subjects, measurements were repeated at the same time of day after 1 week. For each eye, the intersession coefficient of variation was calculated as a measure of reproducibility. We found a coefficient of variation (+/- standard deviation) of 7% +/- 5% for the amplitude of the steady-state pattern electroretinogram, 9.5% +/- 7% for the transient pattern electroretinogram and 1.5% +/- 2% for the latency of the transient pattern electroretinogram. To assess the diurnal variability, during a 15-h period, three pattern electroretinograms were recorded in 10 subjects. No relationship was found between the P50 latency and the time of day. However, the mean amplitude showed a maximum in the morning (9:30 am) and a minimum in the afternoon (2:30 pm). This small effect (about 7%, p < 0.001) was more pronounced for N95 and steady-state amplitudes than for P50 amplitudes (p < 0.01). Diurnal contributions to the pattern electroretinogram ranged between 3% and 10%. We conclude that pattern electroretinogram amplitude reproduces within +/- 10% with a probability of 70%. The effect size of the diurnal variation is similar and might be relevant for longitudinal studies.