Two chemically distinct aldose reductase inhibitors, ponalrestat and tolrestat, were tested against laser Doppler blood flow and conduction deficits in the sciatic nerve of diabetic rats. The effects two months of streptozotocin-induced diabetes and aldose reductase inhibition on body weight, plasma glucose, and nerve sugars and polyols were comparable to those reported previously. Nerve blood flow, reflected by laser Doppler flow measurements, and motor nerve conduction velocity were both significantly less in diabetic than in control animals. Both of these reductions were prevented by ponalrestat, but not tolrestat. Thus, either deficits in laser Doppler blood flow and conduction are not aldose reductase inhibitor-dependent or tolrestat has some other property which offsets the beneficial effects of aldose reductase inhibition. In either case, these data are consistent with the hypothesis that reduced nerve blood flow contributes to conduction deficits in diabetes