Electrophysiological methods like pattern electroretinogram and pattern visual evoked potentials may be the best approach for early glaucoma detection, as they have demonstrated to be sensitive to the ganglion cells functional changes. Reported values of both P50 and N95 pattern electroretinogram components differ widely, a fact perhaps related to a heterogeneous grouping of patients suffering from ocular hypertension and glaucoma, rather than to the type of technique utilized. This study is based on a total of 42 subjects: 14 normals, 16 subjects suffering from ocular hypertension, and 12 patients with glaucoma, with 79 eyes examined. Pattern electroretinogram and pattern visual evoked potentials were used as successive techniques. Setting condition for pattern electroretinogram, such as low temporal frequency (2 Hz), 30 min check size, high contrast (99%) and luminance (93 cd/m2), identified the N95 component as the best index for early glaucoma detection. In the control group N95 mean amplitude was 1.62 muV +/- 0.59 SD and showed almost significant difference with ocular hypertension (p = 0.07) and highly significant difference with the glaucoma group (p < 0.01), with decrement of 58.6% in the glaucoma group. P50 mean amplitude, on the contrary, did not show significant differences among the groups (Newman-Keul test), its reduction in glaucoma being 28%. The mean pattern visual evoked potentials latency was alos highly significant between glaucoma and control groups, but not between ocular hypertension and control groups; the mean amplitude did not show significant differences. A proportion of 26.6% abnormal pattern electroretinogram was found in the group suspected of having glaucoma where conventional methods had proved normal.(ABSTRACT TRUNCATED AT 250 WORDS)