Experimental diabetes in rodents has been successfully treated by implantation of isolated islets using a syngenic system (Lewis rats). It is possible to reverse all diabetic symptoms of the animals and to prevent late complications in kidney, eye and nervous system. Although isolated islets are highly immunogenic in an allogenic system immuno-alteration techniques have been developed and succeeded in longterm survival after culture at low temperature (24 degrees C), UV-irradiation, cryopreservation, pretreatment with Ia-antibodies etc. Islet transplantation in larger animals and in man up to now has been less successful. Although in a few studies longterm survival of canine islets has been observed, other groups were less successful using dogs and pigs in auto- or allo-transplantation. In man there are reports from various institutions during the last fifteen years using adult or fetal islet material. Only in a few instances the patients came off insulin for some weeks or months. The reasons for this failure are probably manifold: low number of islets, impurity, long ischemia time before isolation, transplantation to inappropriate sites, impairment of engraftment in longterm diabetic recipients and recurrence of autoimmunity in transplanted islets. Further studies are necessary to overcome these barriers. Recent observations using a higher number of islets (> 500,000) and new immunosuppressive drugs (FK506) seem to be promising.