This study sought to determine the timecourse of development of reduced nerve laser Doppler flux in experimental diabetes and the effect on this anomaly of insulin treatment. In addition, we aimed to compare nerve laser Doppler flux in streptozotocin-and genetically-diabetic BB rat models. Sciatic nerve laser Doppler flux in diabetic rats was variable during the 2 days following streptozotocin injection; from day 4, when the measurement was 80% of control, fluxes fell steadily and formed a plateau at 40% of control values after 4 weeks of diabetes. In a second study, using rats with 4-week streptozotocin-diabetes, sciatic nerve laser Doppler flux was reduced to 44% of the value measured in control rats. Treatment of a parallel group of diabetic rats with insulin, by sustained release implants, prevented this ischaemia, so that nerve laser Doppler flux was 91% of controls. Nerve Doppler flux in BB rats with 6-week genetic diabetes was 57% of a control (non-diabetic) BB group. There were no differences in mean arterial pressures between control and diabetic rats in any of the studies. Heart rates of control and insulin-treated diabetic animals were higher than those of the untreated diabetic group; in the other studies heart rates of diabetic animals were numerically lower than controls, but not significantly so. These observations suggest that sciatic nerves of rats with short-term diabetes, whether induced with streptozotocin or of genetic origin, are markedly ischaemic and that this ischaemia in streptozotocindiabetes is evident within a week of diabetes onset, forms a plateau after 4 weeks and is maintained for at least 2 months. The findings also indicate that treatment of short-term diabetes with insulin can prevent nerve ischaemia.