In vivo receptor occupancy of psychotropic drugs in brain can be estimated by measuring the tissue radioactivity of the tracer, which binds specifically to the receptor unoccupied by the drugs (back titration method). In this study, the validity of this method was evaluated by computer simulation, using various values for the plasma elimination rate, rate of transport across the blood-brain barrier, rate of receptor association and dissociation for both drug and tracer, and the sampling time. The differential equations based on a nonlinear three-compartment model including a plasma pool, precursor pool, and specific binding pool were solved numerically by the Runge-Kutta-Gill method. The receptor occupancy calculated by this method was close to the true value when the plasma concentration and specific binding fraction of the drug did not change greatly during circulation of the tracer. Although the error in calculated occupancy at 5 min after the tracer administration was smaller than that at 20 min, tracer may not greatly accumulate in brain tissue during the initial 5 min in some situations. Our analysis shows that it is necessary to adequately control the elimination rate of drug from plasma and to allow sufficient time for radioactivity to accumulate in the tissue. Therefore, this method requires previous knowledge of the pharmacokinetic behavior of both the drug and the tracer in plasma and tissue. The operation scheme that we suggest for the accurate measurement of the receptor occupancy in vivo can be used in human studies with positron emission tomography and may be useful for therapeutic drug monitoring.