Perspectives of in vitro dissolution tests in establishing an in vivo/in vitro correlation are considered at three different levels. The first involves the selection of a drug product, or suitable batch(es), for the correlation study. This requires evaluation of the biopharmaceutical properties of the proposed drug from a profile compiled from a set of dissolution tests under different experimental conditions. These characteristics are needed for properly designing the in vivo study and are also referred to when only a poor correlation is obtained. The second is to define the specific in vitro test system either prospectively, by assessing plausibility of in vitro results, or retrospectively, after interpretation of in vivo data. This distinct test system has to be validated both physically and analytically. The third aspect of dissolution testing for correlations is to provide sound data and derived parameters in consideration of statistical aspects (variability) and of a sufficient number of suitable sampling intervals. In vitro models proposed for the special study of the simulation of food effects on bioavailability are discussed but concluded to be as yet not satisfactory for general application.