The present study evaluated whether a microsphere technique could be used for islet blood flow measurements in anesthetized mice. When this was confirmed, we applied the technique in different strains of mice. Approximately 9 x 10(4) microspheres could be given without interfering with mean arterial blood pressure. Mixing of the microspheres with arterial blood was adequate, and the extraction of microspheres in capillary beds was nearly 100%. In NMRI mice whole pancreatic blood flow was estimated to be 0.54 +/- 0.11 ml.min-1.g pancreatic tissue-1 and islet blood flow to be 18 +/- 4 microliters.min-1.g pancreas-1 (n = 12 animals per experiment), whereas corresponding values in lean C57Bl/6 mice were twice as high. In C57Bl/6 mice glucose (3 g/kg iv) doubled islet blood flow without affecting whole pancreatic blood flow, whereas no effect was seen after an equimolar dose of 3-O-methylglucose. In obese-hyperglycemic C57Bl/6 mice, islet blood flow was more than five times higher than in the lean C57Bl/6 mice when expressed as blood flow per gram pancreas. However, when islet blood perfusion was corrected for islet weight, it was lower in the obese than in the lean mice, suggesting an impaired ability in obese mice to increase blood flow in concert with the increased islet mass. This may contribute to the insufficient insulin secretion and resulting hyperglycemia seen in these animals.