A dual study was conducted to assess (1) the long-term antiparkinsonian action of amantadine without levodopa and (2) the advantage of combined amantadine and levodopa over single-drug therapy, including changes in symptom severity when placebo replaces amantadine but levodopa is maintained.Good to excellent results were obtained in 25% of the total pool of 77 patients on amantadine. No decline in therapeutic effect took place during a mean follow-up of 21 months.Thirty-seven patients with considerable residual deficit after single-drug therapy derived improvement from the second drug (levodopa or amantadine). Gains in neurological signs and activities of daily living (ADL) ranged between 50 and 60% and for timed skills close to 25%. Depending on the individual indices, between 64 and 100% of patients improved with the second drug.Placebo instead of amantadine produced deterioration. There was 75% loss in ADL, 45% loss in timed skills and losses in neurological signs exceeded gains produced by two-drug therapy.