A physiologically realistic model of enterohepatic cycling (EHC) which includes separate liver and gallbladder compartments, discontinuous gallbladder emptying and first-order absorption from both an oral formulation and secreted bile (kapo and kab, respectively) has been developed. The effect of EHC on area under the first-moment curve (AUMC) of drug concentration in plasma and on parameters derived from the AUMC was investigated. Unlike AUC, AUMC is dependent on the time and time-course of gallbladder emptying, increasing as the interval between gallbladder emptying increases. Consequently, mean residence time (MRT) is also a time-dependent parameter. Analytical solutions for MRTiv and MRTpo were derived. Mean absorption time (MAT = MRTpo - MRTiv) is also time-dependent, contrary to findings previously published for a model of EHC with a continuous time lag. MAT is also dependent on kapo, kba and the hepatic extraction ratio. The difference between MRTpos for two formulations with unequal kapo values may deviate from the difference in the inverse of their absorption rate constants. Implications for design and interpretation of pharmacokinetic studies include (i) MAT values may be dominated by the time-course of recycling rather than the time-course of the initial absorption, depending on the extent of EHC and (ii) the unpredictable nature of the time of gallbladder emptying will contribute to intrasubject variability in derived parameters during crossover studies. Knowledge of the extent of EHC is invaluable in deciding whether modification of the in vitro release characteristics of an oral formulation will have any effect on the overall time-course of absorption in vivo. Techniques to monitor or control gallbladder emptying may be helpful for reducing variability in pharmacokinetic studies for compounds which are extensively cycled in bile.