Surface tension (ST) measurements of solutions of a number of purified lyso-phosphoryl- (single tail) lipids over a range of times and concentrations, many of which have been previously published, are reanalysed. They provide better evidence for some old ideas and lead to new conclusions. The main results are as follows. In the presence of traces of highly surface-active impurities, information about the c.m.c. of the system is best obtained before the impurity can significantly contaminate the surface. On the other hand enough time must elapse for the main component to approach equilibrium sufficiently as otherwise erroneously high values are obtained. Two agreeing measurements at sufficiently spaced times, such as differing by a factor of two, can insure this condition. In more impure systems it is sometimes possible to distinguish between the rapid ST changes due to the surfactant which, in the bulk of the solution, is in equilibrium with the impurities, and the slow ST changes due to the impurities. The former define a c.m.c. for the system, always lower than the c.m.c. of the pure surfactant, the latter an ST minimum at a concentration which may be still lower by a factor of two or more. The c.m.c. of the pure surfactant may be significantly higher than that of the system because of the scavenging of the impurity by the micelles then existing. The phase state of the impurity (dissolved or precipitated) can be inferred from the extent and rate of ST lowering. The rates of ST lowering, as measured by the inflection point of an ST verus log concentration plot, is found to be largely diffusion controlled as shown by stirring experiments, but not to be inversely proportional to the square root of time as expected in a diffusion-controlled system. Instead an inverse proportionality to time was found in all but one system. © 1990.