A long-term retrospective evaluation (5 years) compares two groups of RDT patients (group 1 on continuous treatment with cellulosic membranes and group 2 with synthetic membranes) regarding survival, general clinical morbidity, and beta 2M-related morbidity. The results showed no significant long-term differences between the groups either for survival or general morbidity despite some differences in biocompatibility. The higher intradialytic removal of beta 2M by synthetic membranes did not lead to a reduction in either pre-dialysis beta 2M values or beta 2M related morbidity. The higher cost of synthetic over cellulosic membranes and the disappointing of many clinical expectations suggest that the use of such membranes, in association with alternative techniques, should take place only according to certain "elective" indications such as old age, diabetes, vascular instability or intradialytic disequilibrium syndrome.