Although mexiletine, an antiarrhythmic with local anesthetic properties, has been reported to relieve discomfort in diabetic neuropathy, its usefulness in the treatment of HIV-related painful peripheral neuropathy (PPN) has not been determined. The tolerance and effectiveness of mexiletine in HIV-related PPN were assessed in 22 patients who were randomized to receive mexiletine (maximum dose, 600 mg/day) or placebo for 6 weeks, followed by the alternative intervention for 6 weeks after a 1-week washout period. The daily pain response was assessed using a visual analogue scale card in 19 patients who received at least 2 weeks of the drug, 16 of whom were crossed-over to receive the alternate agent. No statistically significant difference was found between the mean daily pain scores for patients receiving mexiletine versus placebo, irrespective of the order in which the agents were received. Comparing the mean individual daily pain scores for each phase of study, 5 patients (31%) had significantly less pain while receiving mexiletine compared with their response to placebo, 5 patients (31%) had significantly less pain while receiving placebo, and no difference was noted in 6 patients (38%). Crossover and multivariate analyses for repeated measures showed no apparent difference in the response to mexiletine versus placebo. Dose-limiting adverse events occurred in 39% of those receiving mexiletine, but only 1 patient (5%) discontinued placebo. Mexiletine was only modestly well tolerated despite its relatively brief period of administration, and no evidence was found to support its benefit in HIV-related PPN. Although a first-drug effect was not demonstrated, a powerful placebo effect was seen in some patients.