OBJECTIVE: To evaluate the efficacy and safety of alpha-lipoic acid given intravenously, followed by oral treatment in type 2 diabetic patients with symptomatic polyneuropathy. RESEARCH DESIGN AND METHODS: In a multicenter randomized double-blind placebo-controlled trial (Alpha-Lipoic Acid in Diabetic Neuropathy [ALADIN] III Study), 509 outpatients were randomly assigned to sequential treatment with 600 mg alpha-lipoic acid once daily intravenously for 3 weeks, followed by 600 mg alpha-lipoic acid three times a day orally for 6 months (A-A; n = 167); 600 mg alpha-lipoic acid once daily intravenously for 3 weeks, followed by placebo three times a day orally for 6 months (A-P; n = 174); and placebo once daily intravenously for 3 weeks, followed by placebo three times a day orally for 6 months (P-P; n = 168). Outcome measures included the Total Symptom Score (TSS) for neuropathic symptoms (pain, burning, paresthesias, and numbness) in the feet, and the Neuropathy Impairment Score (NIS). Data analysis was based on the intention to treat. RESULTS: No significant differences between the groups were noted for the demographic variables and the nerve function parameters at baseline. The TSS in the feet decreased from baseline to day 19 (median [range]) by -3.7 (-12.6 to 5.0) points in the group given alpha-lipoic acid intravenously and by -3.0 (-12.3 to 8.0) points in the placebo group (P = 0.447), but the area under curve on a daily basis was significantly smaller in the active as compared with the placebo group (85.6 [0-219] vs. 95.9 [5.5-220]); P = 0.033). After 7 months, the changes in the TSS from baseline were not significantly different between the three groups studied, which could be due to increasing intercenter variability in the TSS during the trial. The NIS decreased after 19 days by -4.34+/-0.35 points (mean +/- SEM) in A-A and A-P and -3.49+/-0.58 points in P-P (P = 0.02 for alpha-lipoic acid versus placebo) and after 7 months by -5.82+/-0.73 points in A-A, -5.76+/-0.69 points in A-P, and -4.37+/-0.83 points in P-P (P = 0.09 for A-A vs. P-P). The rates of adverse events were not different between the groups throughout the study. CONCLUSIONS: These findings indicate that a 3-week intravenous treatment with alpha-lipoic acid, followed by a 6-month oral treatment, had no effect on neuropathic symptoms distinguishable from placebo to a clinically meaningful degree, possibly due to increasing intercenter variability in symptom scoring during the study. However, this treatment was associated with a favorable effect on neuropathic deficits without causing significant adverse reactions. Long-term trials that focus on neuropathic deficits rather than symptoms as the primary criterion of efficacy are needed to see whether oral treatment with alpha-lipoic acid over several years may slow or reverse the progression of diabetic neuropathy.