To test which, if any, of the major milk precursors can elicit a rapid change in the rate of mammary blood flow (MBF) and to define the time course and magnitude of such changes, 4 lactating cows were infused with glucose, amino acids, or triacylglycerol into the external iliac artery feeding one udder half while iliac plasma flow (IPF) was monitored continuously by dye dilution. Adenosine and saline were infused as positive and negative controls, respectively, and insulin was infused to characterize the response to a centrally produced anabolic hormone. To test the roles of cyclooxygenase, NO synthase and ATP-sensitive K (KATP) channels in nutrient-mediated changes in blood flow, their respective inhibitors-indomethacin, Nω-nitro-L-arginine methyl ester hydrochloride (L-NAME), and glibenclamide-were infused simultaneously with glucose. Each day, 1 infusate was given twice to each cow, over a 20-min period each time, separated by a 20-min washout period. In addition, each treatment protocol was administered on 2 separate days. A 73% increase in IPF during adenosine infusion showed that the mammary vasodilatory response was quadratic in time, with most changes occurring in the first 5min. Glucose infusion decreased IPF by 9% in a quadratic manner, most rapidly in the first 5min, indicating that a feedback mechanism of local blood flow control, likely through adenosine release, was operative in the mammary vasculature. Amino acid infusion increased IPF 9% in a linear manner, suggesting that mammary ATP utilization was stimulated more than ATP production. This could reflect a stimulation of protein synthesis. Triacylglycerol only tended to decrease IPF and insulin did not affect IPF. A lack of IPF response to glibenclamide indicates that KATP channels are not involved in MBF regulation. Indomethacin and L-NAME both depressed IPF. In the presence of indomethacin, glucose infusion caused a quadratic 9% increase in IPF. Indomethacin is an inhibitor of mitochondrial function, so the glucose-induced increase in IPF was interpreted as feedback on mammary adenosine release from an anabolic response to glucose. Because NO synthase was not inhibited during indomethacin infusion, the feedback system is postulated to act through endothelial NO synthase. In the presence of L-NAME, glucose infusion had no effect on IPF, indicating that endothelial cyclooxygenase is not involved in glucose-induced changes in MBF.