The critical micellar concentration (cmc) values of some mixed systems containing two bile salts were determined by a maximum pressure bubble method and compared with those derived from a theoretical model developed for nonionic surfactants to assess the applicability of this model to such systems. Some assumptions on which the presumed validity of this model was based are discussed. The following binary mixtures were investigated: sodium chenodeoxycholate with cholate, ursocholate and ursodeoxycholate, either unconjugated or conjugated with taurine and glycine at different mole fractions (0, 0.25, 0.5, 0.75, 1) in 0.15 M NaCl. For these mixtures, experimentally determined data were in good agreement with values predicted by the theoretical model: both the cmc and the surface tension at this concentration of the mixtures were intermediate between those of the two pure bile salts; also, as the total bile salt concentration increased, the mixed micelles became enriched with the bile salt having the highest cmc, whereas the total monomer activity, determined by a potentiometric method employing a bile salt-selective electrode, increased only slightly. To test this model in an in vitro system, surface tension was also measured in ox bile samples that were enriched by 50% with sodium ursodeoxycholate, chenodeoxycholate, or their taurine amidates. The cmc and the surface tension at this concentration of the artificial bile increased when enriched with a bile salt with a cmc higher than that of endogenous salts (e.g. ursodeoxycholate versus taurocholate), whereas the reverse occurred for mixtures enriched with a bile salt with a lower cmc, such as chenodeoxycholate.(ABSTRACT TRUNCATED AT 250 WORDS)