We explored the use of top-down Physiologically Based Finite Time Pharmacokinetic (PBFTPK) models in early drug development. The Absorption number, An of drugs was re-defined as the ratio of residence time of drug in the gastrointestinal tract and the half duration of its absorption, τ. Using estimates for τ of several drugs derived from the fittings of PBFTPK models to literature experimental data and the mean intestinal transit time, 199 min, we calculated the corresponding An values and discussed them in terms of their BCS classification. Analysis of amoxycillin data using PBFTPK models with passive absorption or carrier mediated transport was also carried out. Low τ values result in high An estimates. The mean duration of absorption for Class I, II and III drugs was found to be equal to 1.96 ± 1.91, 3.63±8.71, 1.81±1.02 h, respectively. The shorter duration of absorption for Class I and III drugs was attributed to their high solubility. The analysis of PBFTPK model fittings revealed that amoxycillin absorption follows carrier mediated transport since both the peak blood concentration and AUC values are increasing as a function of dose nonlinearly following the same pattern. The definition of An in terms of the duration of absorption is physiologically sound since the high blood flow rate in the portal vein imposes sink conditions, namely zero order kinetics in oral drug absorption. The use of the PBFTPK models in early drug development provides meaningful estimates for the number of absorption stages and their duration as well as the corresponding input rates.