This study investigated the effects of the traditional herbal medicine, Keishi-ka-jutsubu-to (KJT) on insulin action in vivo and insulin signaling in skeletal muscle in STZ-induced diabetes. Rats were divided into single and 7-days oral administration groups. Euglycemic clamp (insulin infusion rates: 3 and 30 mU/kg/min) was used in awaked rats and the insulin signaling in skeletal muscle was evaluated. At low-dose insulin infusion, the decreased metabolic clearance rates of glucose (MCR) in diabetic rats were improved by a single and 7-days administration of KJT (800 mg/kg BW, p.o.; acute effect: 6.7 +/- 0.6 vs. 12.3 +/- 1.2, and 7-days effect: 6.3 +/- 0.5 vs. 13.9 +/- 1.0 ml/kg/min, P<0.001, respectively). During high-dose insulin infusion, the MCR was increased in 7-days KJT treated diabetes compared with saline diabetes, but, these changes were not observed after a single KJT treatment. About 90% of the increasing effect in MCR induced by the 7-days KJT treatment was blocked by L-NMMA. However, no further additive effects were seen in KJT + SNP treatment. IRbeta protein increase and decreased IRS-1 protein expression in diabetes were significantly improved by KJT treatment. KJT had no effect on the GLUT4 protein content. The increased tyrosine phosphorylation level of IRbeta, IRS-1, and IRS-1 associated with PI 3-kinase were significantly inhibited in KJT treated diabetes. The present study suggests that the improvement of impaired insulin action in STZ-diabetes by administration of KJT may be due, at least in part, to enhanced insulin signaling, which may be involved with production of nitric oxide (NO).