The aim of this study was to determine whether the beta-adrenoceptor receptor density (Bmax) and the ligand affinity (KD) of (S)-[11C]CGP 12388 for the beta-adrenoceptor receptor could be determined using full tracer kinetic modelling of the transport of the ligand and its interaction with the receptor. This approach minimises the a priori assumptions and may thus serve as a gold standard to validate other simplified methods. Dynamic positron emission tomography (PET) data were acquired in six healthy subjects during 60 min. Three different injection protocols were applied, each consisting of three injections with varying SAs: high specific activity (SA), low SA or unlabelled ligand only. Arterial blood samples were collected via a cannula in the radial artery. Time-activity data in myocardial tissue were obtained using regions of interest (ROIs) on short-axis planes. All data were analysed with a two-tissue compartment, six-parameter (K1, k2, k(on), k(off), Bmax, F(bv)) model that relies on explicit compartments for describing the kinetics of both labelled and unlabelled radioligand. Time-activity curves showed that unlabelled ligand could displace the radioligand from the receptor. This resulted in increased radioactivity levels in plasma. Modelling results yielded Bmax values of 9.74+/-1.80 nM and a KD of 0.58+/-0.22 nM, assuming a reaction volume of 0.15. In addition, parametric polar images of Bmax could be calculated. The protocol with injections of high SA, low SA, and unlabelled ligand, respectively, was found to be the most sensitive to parameter changes. We conclude that with tracer kinetic modelling of (S)-[11C]CGP 12388, the beta-adrenoceptor density in the human heart can accurately be obtained in vivo. This approach may thus serve as a gold standard.