Abstract Equations are given for the dissolution of a drug into an aqueous phase overlayered with a lipid phase, in which back transfer from the lipid phase is assumed possible. Consideration of both the presence and absence of back transfer, assuming in each that the amount of drug in the aqueous phase either approaches saturation or remains negligible, results in four equations for following the amount of drug in the lipid phase versus time. Three of the four equations are identical in form, predicting that after a brief lag time, a plot of the amount of drug in the lipid phase versus time will become linear. However, only in the instance for which no back transfer occurs and for which the buildup of drug in the aqueous phase is negligible (Case IV) is the slope of the line equal to the zero-order dissolution-rate constant, k 0 . In the other cases, the slope represents a fraction of k 0 , depending upon the relative rates of dissolution, partitioning, and back transfer. Thus, it becomes important to be completely certain that the conditions under which dissolution studies into a „perfect sink” are carried out do actually meet the requirements of Case IV.