OBJECTIVE To evaluate motion-evoked brain potentials for glaucoma diagnosis. METHODS Stripe patterns were presented in Maxwellian view under different stimulus conditions not combined in a factorial design. Spatial frequencies of 0.33 and 0.88 cycles/degree, with speeds of 10 and 5.9 degrees/second and contrasts of 0.04 and 0.93 were used. A 32 degrees whole field and a peripheral 32 degrees-27 degrees annular stimulus were used. Duration of motion was 200 milliseconds, and the interstimulus interval was 1,800 milliseconds. Recordings were obtained from Oz and P3. Thirty-four healthy patients, 12 glaucoma suspects, and 26 patients with open-angle glaucoma were tested. RESULTS Normal response amplitudes decrease with age only under low contrast conditions, whereas response peak times increase under most conditions. Normal responses are much larger at P3 than at Oz, whereas in open-angle glaucoma, much less difference is seen. In these patients, the response amplitude at P3 is significantly reduced under all conditions, whereas a delay in peak time is less pronounced. A small but significant negative correlation (r = -0.44, P < 0.05) between response amplitude and mean perimetric defect was observed only with the annular stimulation. At a specificity of 90%, a sensitivity of approximately 76.7% for the low contrast and low spatial frequency condition was observed. CONCLUSIONS Motion-evoked responses recorded at P3 are altered in open-angle glaucoma and thus can be useful as an additional test in glaucoma diagnosis.