To compare the kinetics of praehepatic and of posthepatic administered insulin, short term insulin deprived diabetic dogs were sequentially injected with 200 mU/kg of a monocomponent porcine insulin using either the intravenous, portal, subcutaneous or peritoneal route. After peritoneal insulin was applied, the peripheral plasma insulin levels increased immediately, their maxima were in the same range as after subcutaneous injection but the duration of elevation was shorter. There were portal-peripheral insulin-quotients greater than 1 after peritoneal and portal insulin administration but quotients less than 1 after subcutaneous and intravenous application. The time constant of insulin elimination was identical regardless of whether the praehepatic or the posthepatic route was used for application. The effectiveness of the administered insulin dose on blood glucose was found to be dependent on the posthepatic elevation of plasma insulin and its duration. The decrease in glycemia was initially identical in all tests but, on the whole, it was smaller after the two intravascular routes were used because of the shorter duration of elevated insulin levels. It is concluded that in an optimized management of insulin-dependent diabetes, the regime (doses and intervals or algorithms) must be adapted to the pharmacokinetic implications of the employed route of application.