We investigated the relationship between levodopa plasma concentration and the tapping effect, after a standard oral levodopa test, by kinetic-dynamic modeling in 40 parkinsonian patients with stable or fluctuating response to levodopa, and found no difference in levodopa plasma pharmacokinetics between stable and fluctuating patients. Conversely, levodopa equilibration half-life between plasma and effect-site concentration was five-fold shorter on average in fluctuating patients. Overall, levodopa equilibration half-life highly correlated with the duration of tapping response and provided a reliable quantitative index of central mechanisms that affect the length of clinical effect. Individual fitting of tapping measures to modeled drug effect-site concentrations by sigmoid Emax model revealed that fluctuating patients required almost two-fold higher levodopa concentrations (EC50) to elicit almost the same motor response (Emax). These findings suggest that shortening of levodopa clinical effect may be accompanied by a reduced drug affinity for the nigrostriatal dopaminergic system (EC50), with no change in its intrinsic activity (Emax).