The present study used recordings of visual potentials evoked by pattern reversal (VEPs) to investigate the central effects of three drugs used in migraine prophylaxis: the calcium channel blocker nifedipine, the beta-1-selective blocker metoprolol, and the nonselective beta adrenoreceptor blocker propranolol. The study involved 58 patients with common or classical migraine who were treated in a double-blind randomized study over a period of 7 months, while the effectiveness of prophylactic treatment was recorded in headache diaries that were subjected to time series analysis. VEPs were recorded at the beginning of a 2-month baseline period without treatment, after 4 months of treatment, and at the end of a 3-month washout period. At baseline, migraine patients had significantly higher VEP amplitudes and longer latencies than did a group of 87 healthy control subjects. Patients were separated by statistical analysis into responders and nonresponders to each prophylactic treatment. Nifedipine had no effects on the frequency, intensity, and duration of migraine attacks, nor on amplitude and latency of the VEPs. In contrast, the use of beta blockers resulted in a significant decrease in VEP amplitude, both in responders and nonresponders, whereas VEP latency remained unchanged. VEP amplitudes returned to the initial values at follow-up in the nonresponders, but stayed at lower levels in responders. Beta blockers thus appear to have a significant effect on the increased excitability of the visual system in patients with migraine, although their action is not directly related to their reduction of migraine frequency.