Angiotensin II (ANG II) was infused intravenously within an apparent physiological dose range of 10-200 ng.kg-1.min-1 to induce a drinking response in rats. To determine the plasma ANG II level at the moment of drinking, blood samples were collected from the femoral artery at the onset of the drinking response. Control blood samples were obtained in a similar way before infusion. The lowest dose of ANG II did not evoke drinking. Variable water intake in response to 25, 50, and 75 ng ANG II.kg-1.min-1 after a relatively long latency to drink (40-65 min) was observed. All animals showed a drinking response when 100 ng ANG II.kg-1.min-1 or higher doses of ANG II were infused. The latencies to drink were inversely correlated to the dose of ANG II and were as short as 8 min with 200 ng ANG II.kg-1.min-1. Measurement of ANG II in the plasma of drinking rats showed that all concentrations were similar with an average of 458 +/- 58.1 pg/ml. This dipsogenic plasma ANG II level is equivalent to plasma ANG II after 48 h of dehydration. The results show that drinking in response to exogenous intravenous ANG II requires a threshold level to be reached that is equivalent to levels produced by the endogenous renin-angiotensin system when dehydration is prolonged. This suggests that intravenous ANG II is not involved in moment-to-moment fluid homeostasis but operates only when dehydration is severe.