The enzymatic conversion of in aggregate-forming Substrate was kinetically analyzed and a model was applied for the prediction Of reaction-time Courses. An L-rhamnose molecule from a di-rhamnolipid is cleaved by Naringinase from Penicillium decumbens leading to a mono-rhamnolipid. Optimal reaction rates were found when both, substrate and product build large co-aggregates in a slightly acidic aqueous phase. On the other hand, reaction rates were independent of initial di-rhamnolipid concentration and this was interpreted by assuming that the reaction Occurs in the aqueous phase according to Michaelis-Menten kinetics in combination with competitive L-rhamnose inhibition. Rhamnolipids were therefore assumed to be highly concentrated in aggregates, a second liquid phase, whereas diffusive rhamnolipid transport from and to the aqueous phase occurs due to the enzymatic reaction. Furthermore, ideal surfactant mixing between di-and mono-rhamnolipid was assumed for interpretation of the negative effect of the last oil the reaction rate. A model was created that describes the system accordingly. The comparison of the experimental data, were in excellent agreement with the predicted Values, The findings of this study may beneficially be adopted for any bioconversion involving aggregate-forming substrate and/or product being catalyzed by hydrophilic enzymes. Biotechnol. Bioeng, 2009;102: 9-19. (C) 2008 Wiley Periodicals, Inc.