A systematic approach to the phenomenon of surfactant-dependent release of liposomal contents has been attempted. A variety of methods have been comparatively studied. The influence of the size of the entrapped molecule, nature of the surfactant, composition of bilayers and sonication of liposomes have been considered separately. In order to compare different results, a parameter has been defined, R50, as the phospholipid/surfactant mole ratio producing 50% release of the entrapped solute. This parameter appears to be, to a large extent, independent of time and liposome concentration. Surfactant-induced release of liposomal contents does not occur as a result of breakdown of phospholipid bilayers, but is rather a different phenomenon, occurring at detergent concentrations substantially lower (2-5 times) than solubilization. The required amount of surfactant appears to increase with the size of the entrapped solute. R50 depends clearly on the nature of the soluble amphiphile, but there is no obvious relationship with its critical micellar concentration. Liberation of vesicle content also depends on bilayer composition: phospholipids have various effects on the stability of the membrane, while the hydrophobic peptide, gramicidin A, appears to have little influence. Cholesterol is interesting, since at equimolar proportions with phosphatidylcholine, it decreases the stability of bilayer towards Triton X-100, while increasing it in the presence of cholate. Sonication also exerts an influence on the surfactant-dependent release of vesicle contents; it appears to decrease the bilayer stability, so that lower detergent concentrations are required to liberate the entrapped solutes. Finally, it should be noted that, although the decrease in self-quenching of 6-carboxyfluorescein is a convenient method for the study of solute liberation, glucose release, as detected by enzymatic methods, may be more reliable for accurate measurements.