The purpose of this study was to continue our effort to determine how enzyme-transporter coupling affect disposition of flavonoids. The rat intestinal perfusion and Caco-2 cell models were used together with relevant microsomes. In perfusion model, isoflavone (i.e., formononetin and biochanin A) absorption and subsequent excretion of its metabolites were always site-dependent. Maximal amounts of intestinal and biliary conjugates excreted per 30 min were 31 and 51 nmol for formononetin, more than that for pure biochanin A ( 12 and 20 nmol). When a standardized red clover extract ( biochanin A/formononetin = 10: 7) was used, the results indicated that more metabolites of biochanin A than formononetin were found in the perfusate (36.9 versus 22.8 nmol) and bile ( 78 versus 51 nmol). In metabolism studies, rat intestinal and liver microsomes always glucuronidated biochanin A faster ( p < 0.05) than formononetin, whereas intestinal microsomes glucuronidated both isoflavones faster ( p < 0.05) than liver microsomes. However, rapid metabolism in the microsomes did not translate into more efficient excretion in either the rat perfusion model as shown previously or in the Caco-2 model. In the Caco-2 model, both isoflavones were rapidly absorbed, efficiently conjugated, and the conjugates excreted apically and basolaterally. More formononetin conjugates were excreted than biochanin A when used alone, but much more biochanin A conjugates were found when using the isoflavone mixture. In conclusion, efficiency of enzyme-transporter coupling controls the amounts of metabolites excreted by the intestine and liver and determines the relative contribution of enteric and enterohepatic recycling to the in vivo disposition of isoflavones.