The acute dose-response profile of a standard oral levodopa dose was followed, over a maximum 8-h period, in 13 patients with and 10 patients without motor fluctuations using a battery of motor quantitative tests (tapping and walking speed, and multiple choice reaction and movement times). Thirteen age-matched normal controls performed tapping and psychomotor tests, at the same time intervals, over a 4-h period. Tapping test and movement times proved significantly impaired in all patients and were the best indicator of levodopa effect, while walking speed and reaction times were apparently of less value, except in severely affected patients. The duration of the levodopa antiparkinsonian effect differed markedly between the two groups, since fluctuating patients returned to prelevodopa dose values within 4 h (mean ± SEM: 203 ± 16 min), while in the stable group motor scores remained significantly higher than baseline values up to at least 7 h postdose. The magnitude of the effect was similar in the two groups, but response was complicated by mild to severe dyskinesias in 9 of 13 fluctuating subjects. The pharmacokinetic parameters of levodopa were almost identical in the two groups. Our data add further weight to the hypothesis that cerebral pharmacokinetic or pharmacodynamic factors are responsible for motor fluctuations. Oral levodopa doses coupled with objective tests of motor performance may prove a practical clinical tool to assess and optimize the relationship between drug dose and therapeutic effect.