Objective. We performed a systematic review of the literature to evaluate the effects of alpha lipoic acid for symptomatic peripheral neuropathy in patients with diabetes mellitus. Research design and methods. The databases MEDLINE and EMBASE were searched using the key words "lipoic acid", "thioctic acid", "diabet∗", and the MeSH-terms "thioctic acid" and "diabetes mellitus". Randomised controlled trials using the TSS score as the outcome measure were selected and assessed for their methodological quality. Study selection and quality assessment were performed independently by three observers. Results. Overall, the pooled standardized mean difference estimated from all trials revealed a reduction in TSS scores of -2.26 (CI: -3.12 to -1.41; P = 0.00001) in favour of alpha lipoic acid administration. Subgroup analyses of oral administration (-1.78 CI: -2.45 to -1.10; P = 0.00001) and intravenous administration (-2.81 CI: -4.16 to -1.46; P = 0.0001) confirmed the robustness of the overall result. Conclusions. When given intravenously at a dosage of 600 mg/day over a period of 3 weeks, alpha lipoic acid leads to a significant and clinically relevant reduction in neuropathic pain (grade of recommendation A). It is unclear if the significant improvements seen after 3-5 weeks of oral administration at a dosage of >600 mg/day are clinically relevant.