The oxidant stress theory of Parkinson's disease (PD) hypothesizes that levodopa treatment may be potentially harmful and this is supported by studies demonstrating levodopa toxicity to cultured dopaminergic neurons. These in vitro experiments, however, lack the physiologic protective mechanisms present in vivo. Oxyradical damage to cell membranes liberates malondialdehyde, which we measured in the serum of 27 PD patients just before and after levodopa (with carbidopa) administration. We also measured plasma products of the two routes by which levodopa potentially generated oxyradicals: (1) 5-S-cysteinyl-dopa (derived from levodopa autoxidation), and (2) 3,4-dihydroxyphenylacetic acid (DOPAC), produced by monoamine oxidase (MAO) metabolism of dopamine. Following levodopa/carbidopa administration, both of these plasma products were markedly increased; however, the mean serum malondialdehyde concentration was unchanged and remained similar to the normal control group (N=15) value. Chronic treatment with the MAO-B inhibitor, deprenyl (N=16), was not associated with any differences in serum malondialdehyde or plasma 5-S-cysteinyl-dopa concentrations compared with those not treated with deprenyl (N=11). The post-levodopa rise of plasma DOPAC was only slightly attenuated with deprenyl therapy, consistent with a predominant MAO-A effect in the circulation and peripheral organs. Thus, in contrast to in vitro studies, we did not detect evidence of oxidative damage in the circulation following levodopa administration, despite marked increase in the products of dopamine oxidative metabolism.