Basic aspects and recent advances in the understanding of the pharmacological mechanism of action of the clinically most used antiparkinson drugs are reviewed. Recent human and animal biochemical investigations clearly confirm and extend previous findings indicating that benserazide is much more potent than carbidopa as peripheral decarboxylase inhibitor. L-DOPA in combination with benserazide or carbidopa constitutes the best available therapy for Parkinson's disease (PD). To reduce peaks and rapid fluctuations of L-DOPA plasma levels (possibly responsible for peak-dose dyskinesias and end-of-dose deterioration) a slow-release formulation of L-DOPA in combination with benserazide or with benserazide plus catechol-O-methyltransferase inhibitors should be developed. In parkinsonian patients under long-term L-DOPA therapy monoamine oxidase inhibitors type B (MAO-B) e.g. (-)deprenyl and direct dopamine receptor agonists (bromocriptine, lisuride, pergolide etc.), due to their L-DOPA-sparing effects, alleviate in some cases L-DOPA-induced side-effects e.g. dyskinesias and on-off phenomena. However, since (-)deprenyl, due to its metabolism to (-)methamphetamine and (-)amphetamine, seem to have indirect sympathomimetic activity, new selective MAO-B inhibitors devoid of indirect sympathomimetic effects should be tested clinically to assess the functional role of pure MAO-B inhibition in the therapy of PD. The auxiliary therapy with direct dopamine receptor agonists of the D-2 subtype represents another valid approach which should be further investigated in order to find novel dopamine agonists, less expensive than bromocriptine, and strictly selective for D-2 receptor sites.