Albino Oxford rats made diabetic with 75 mg/kg streptozotocin were intraperitoneally transplanted with 2500-2900 alginate-polylysine microencapsulated Lewis islets (n = 9, total islet tissue volume 8.0-11.0 microliters), or a similar volume of non-encapsulated Lewis islets (n = 5). All rats with microencapsulated islets became normoglycaemic, and remained normoglycaemic for 5-16 weeks. In rats with non-encapsulated islet grafts, only a temporary decrease in blood glucose was observed, and all were again severely hyperglycaemic at 1 week after implantation. At 5-6 weeks after transplantation, glucose tolerance in rats with microencapsulated islets was tested by intravenous glucose infusion (10 mg/min over 20 min) and test meal administration (n = 4). During glucose infusion, maximum glucose levels were 13.0 +/- 0.4 mmol/l in rats with microcapsules and 8.9 +/- 0.4 mmol/l in healthy control rats (p less than 0.01). Concomitant maximum plasma insulin levels were 215 +/- 17 pmol/l in rats with microcapsules and 715 +/- 85 pmol/l in controls (p less than 0.001). After the test meal, maximum blood glucose was 10.6 +/- 0.9 mmol/l in rats with microcapsules and 6.2 +/- 0.1 mmol/l in controls (p less than 0.001), with concomitant maximum plasma insulin levels of 247 +/- 11 pmol/l and 586 +/- 59 pmol/l, respectively (p less than 0.001). In conclusion, although the glucose tolerance is impaired and plasma insulin responses to intravenous glucose-load and test-meal are reduced, the alginate-polylysine membrane does provide adequate immunoisolation for the prolongation of allograft survival, resulting in prolonged normoglycaemia in streptozotocin diabetic rats.