Twenty-seven patients with Parkinson's disease participated in a double-blind crossover trial of L-dopa and amantadine. Each drug was given for six weeks, and six weeks were allowed to elapse between the two periods of treatment. Marked improvement occurred in patients given L-dopa first, whereas no clinical effect was observed in those patients treated first with amantadine. L-dopa was less beneficial to patients who had previously received amantadine, whereas amantadine became effective in patients who had previously taken L-dopa. The effect of pre-treatment with one drug on the therapeutic effectiveness of the other drug is reflected in changes of homovanillic acid levels in the cerebrospinal fluid.