The objective of this workshop was to review and discuss the debate on neurotoxicity of levodopa in the treatment of Parkinson's disease (PD) with consideration of preclinical and clinical findings. We concluded that in particular preclinical outcomes of in vitro models of neurodegeneration describe neurotoxic effects of levodopa, whereas trials in animal models provided controversial results. To date, clinical trials in PD patients showed no convincing proof of direct neurotoxic effects of levodopa on progression of neurodegeneration with various applied functional imaging techniques particularly with specific radiotracers for nigral dopaminergic neurotransmission. However, the controversy on neurotoxicity of levodopa only partially considered indirect mechanisms, i. e. levodopa-associated homocysteine elevation. But there is accumulating evidence that this long-term side effect of chronic levodopa administration dose dependently individually contributes to progression of neurodegeneration due to increased release of neurotoxins, induction of oxidative stress and mitochondrial dysfunction according to results of in vitro and animal trials and to at least peripheral neuronal degeneration and increased risk for onset of atherosclerosis-related disorders according to clinical trials in PD patients. From this point of view we demand that future research on the efficacy and putative neurotoxicity of antiparkinsonian compounds should also consider putative toxic long-term effects of drug administration and should look for putative peripheral biomarkers and individual, environmental or nutritative risk factors in order to establish a preventive therapy, i. e. folic acid administration in the case of levodopa-associated homocysteine elevation.