In six patients with Parkinson's disease exhibiting severe "on-off" phenomena, a 200-mg intravenous bolus of either L-DOPA or of its methyl ester were equally effective in reversing motor deficits, although the duration of action of the methyl ester was shorter. There were no marked differences in pharmacokinetic parameters for L-DOPA plasma levels after administration of L-DOPA and the methyl ester. In three patients, optimal infusion rates for the maintenance of mobility were established for L-DOPA and L-DOPA methyl ester. Both drugs were able to maintain patients "on" throughout a 12-h infusion period. However, on average the optimal infusion rate of L-DOPA methyl ester was 2.7 times greater than that for L-DOPA. There was no marked difference in the plasma levels of L-DOPA achieved, but 3-O-methyl DOPA levels increased more after infusion of L-DOPA methyl ester than after infusion of L-DOPA itself. The half-life of elimination and volume of distribution of L-DOPA formed from the methyl ester were markedly increased compared with values obtained after either an intravenous bolus of methyl ester or after an intravenous infusion of L-DOPA itself. An intravenous bolus of L-DOPA methyl ester produces an equivalent magnitude of clinical response to the same dose of L-DOPA. However, higher optimal infusion rates of methyl ester than L-DOPA are required to produce continuous effect. The pharmacokinetic handling of L-DOPA methyl ester given by intravenous infusion may differ from that of L-DOPA when given by the same route.