Apomorphine (APO) stimulates growth hormone (GH) release via dopamine D2 receptors (DRD2). There is no specific study assessing the relationship between APO pharmacokinetic (PK) and the pharmacodynamic (PD) response e.g. GH release. The objective of the study is the PK-PD modelling of APO in healthy subjects. This is a randomized crossover study with s.c. administration of 5, 10, and 20 micro g/kg of APO in 18 healthy subjects. APO concentrations were modelled according to both a bi-compartmental model with zero-order absorption and a bi-compartmental model with first-order absorption. PK-PD relationship was modelled in accordance with the Emax Hill equation using plasma concentrations of APO calculated according to the bi-compartmental model with zero-order absorption. Modelled parameters were very similar to the experimental parameters. PK of APO was linear and there was no significant difference between the tested doses for AUC0--> infinity and Cmax (normalised to the dose 1 micro g/kg), t1/2alpha and t1/2beta. These parameters expressed as mean (CV%: SD/mean) were: 17.2 (26.9) ng/mL.min, 0.26 (33.3) ng/mL, 17.1 (54.2) and 45.2 (20.6) min, respectively (n = 53). An anticlockwise hysteresis loop (effect function of APO plasma concentration) appeared for each dose and each subject. The predicted and measured GH concentrations for all subjects and times were similar whatever the dose (P > 0.27). Emax values were 246 (121), 180 (107), 205 (139) ng/mL, respectively, and EC50 were 0.98 (48.1), 1.70 (62.3), 3.67 (65.2) ng/mL, respectively at dose 5, 10, and 20 micro g/kg (P < 10-4). APO and GH concentrations were predicted with good accuracy using bi-compartmental with zero-order absorption PK model and sigmoid Emax PD model, respectively.