Computer simulations were performed to determine if the threshold dose of an infused drug (rather than the drug concentration in the biophase at onset of action) can be a suitable index for pharmacodynamic investigations as proposed by others. A two-compartment pharmacokinetic model with drug elimination from the central compartment was used for the simulations. Drug was administered into the central compartment by a constant-rate infusion, and concentrations in the central and peripheral compartments were calculated as a function of time. The pharmacologic effect was assumed to be reversible and to occur at a defined concentration (the effective concentration) in one or the other compartment. The dose required to produce an effective concentration (threshold dose) was determined as a function of infusion rate. The relationship between infusion rate and the dose required to produce an effective concentration in the peripheral compartment was found to be affected by drug distribution and elimination kinetics and by the effective concentration. The infusion rate–dose relationship showed a dose minimum at an infusion rate which others have designated as the "optimal dose rate” and have used for pharmacodynamic studies. No such minimum occurred for pharmacologic effects associated directly with drug concentrations in the central compartment. Since optimum dose rate and threshold dose are affected by both pharmacokinetic and pharmacodynamic alterations, it is concluded that this method (which avoids determination of drug concentrations) is not generally suitable for quantitative pharmacodynamic investigations.