In vitro/in vivo correlations and the parameters that should be correlated have now been well defined in the USP and FDA working group proposals. Among the three levels of correlation defined, Level A is the most interesting one. In this paper, the Level A correlation is studied from a mathematical and a biopharmaceutical point of view. The conditions that must be met before attempting to establish correlations and the place of in vitro/in vivo correlation in modified release drug dosage form development are discussed. To establish in vitro/in vivo correlations requires a strict development methodology in order to obtain them in the most favourable conditions. In vitro/in vivo correlations should be sought as early as possible during the dosage form development (a priori correlations). If the formulation has been undertaken first and correlations sought on the finished product from subsequent in vitro tests, the predictive power of these correlations (a posteriori correlations) is thus limited and they require additional validation.