BACKGROUND Streptozotocin (STZ) is commonly used to induce diabetes in experimental animal models, but not without accompanying cytotoxic effects. This study was undertaken to (1) determine an optimal dose and administration route of STZ to induce diabetic nephropathy in wild-type mice but without the concurrent acute renal injury resulting from cytotoxic effects of STZ and (2) evaluate the pattern of tubular injury and interstitial inflammation in this model. METHODS Male Balb/c mice received either (1) STZ (225 mg/kg by intraperitoneal injection.); or (2) two doses of STZ 5 days apart (150 mg/150 mg/kg; 75 mg/150 mg/kg; 75 mg/75 mg/kg; and 100 mg/100 mg/kg by intravenous injection). Another strain of mice, C57BL/6J, also received STZ (200 mg/kg intravenously or intraperitoneally). Renal function and histology were examined at weeks 1, 2, 4, and 8 after induction of diabetes. In initial optimization studies, animals were sacrificed at week 1 or week 2 and histology examined for acute renal injury. RESULTS Following a single intraperitoneal injection of 225 mg/kg of STZ, only two thirds of animals developed hyperglycemia, yet the model was associated with focal areas of acute tubular necrosis (ATN) at week 2. ATN was also observed in C57BL/6J mice given a single intravenous or intraperitoneal dose of STZ (200 mg/kg), at week 2 post-diabetes. At an optimal diabetogenic dose and route (75 mg/150 mg/kg by intravenous injection 5 days apart), all mice developed diabetes and no ATN was observed histologically. However, even with this regimen, glomerular filtration rate (GFR) was significantly impaired from week 2. This regimen was accompanied by progressive histologic changes, including tubular and glomerular hypertrophy, mesangial area expansion, as well as interstitial macrophage, CD4+ and CD8+ T-cell accumulation. CONCLUSIONS By careful optimization of STZ dose, a stable and reproducible diabetic murine model was established. However, even in this optimized model, renal functional impairment was observed. The frequency of ATN and functional impairment casts doubt on conclusions about experimental diabetic nephropathy drawn from reports in which ATN has not been excluded rigorously.