It soon becomes apparent to those who use cardiac glycosides clinically that the relation between the amount of drug administered and the therapeutic effect is highly variable. The correlation of dose and effect of the drug is influenced by a number of factors including bioavailability, completeness of absorption, apparent volume of distribution, rate of elimination, degree of plasma protein binding, diffusion and active transport to the site of action, and local factors at the side of action (Koch-Weser, 1972;Sjoqvist and Bertilsson, 1973). KochWeser (1972) effect depended on the drug itself and not on metabolites, (2) protein binding was minimal, (3) the concentration/response graph followed a known and readily definable form such as the linear or monoexponential, and (4) the functional state of the drug receptor responded predictably and instantaneously, or nearly so, to changes in plasma concentration. The first condition is satisfied by digoxin but the second and third conditions are only partially satisfied, and the fourth is not. The importance of the second and third points is crucial and may be illustrated by recourse to the conceptual device of mathematical modelling. The simplest practical model to consider in this situation is the two-compartment model first comprehensively described by Teorell in 1937(Vere, 1972. Administered drug rapidly enters a 'space' where thorough mixing soon occurs (first compartment f blood and some part of well-perfused tissues) and thereafter exchanges with a 'space' more remote from the outside world by virtue of relatively slower ingress and egress of drug across the intercompartmental boundary (second compartment t cells and most drug receptor sites). Intercompartmental transfer is treated as a two-way flux, and absorption and elimination of the drug as one-way fluxes according to the Fick equations for simple diffusion, a valid approach simply because most drugs are effectively non-electric in their transfer behaviour. Teorell showed that very soon after full absorption of a drug, all the terms in the mathematical description of the fluxes across the various boundaries become vanishingly small except for the one-way flux of drug removal from tissue. This leaves a single exponential for removal of drug from the blood and so, unless blood volume alters, plasma concentrations of the drug should follow this exponential form (Vere, 1972).