A new method for measurement at short intervals of uptake or production of metabolites by a perfused heart has been applied to the study of the time-course of glucose utilization. In the presence of high concentrations of insulin (100 mU/ml.) glucose utilization was constant from the beginning of perfusion at all concentrations of glucose. In the absence of added insulin, when the perfusate glucose concentration was greater than 1 mm, glucose utilization decreased by more than 20% to reach a constant rate after up to one hour of perfusion. At glucose concentrations below 0 5 ml, glucose utilization without insulin was constant, or nearly so, throughout perfusion. These results suggest that endogenous insulin, which may be relatively ineffective in promoting glucose utilization at low glucose concentrations, influences the properties of the isolated heart longer than is generally supposed.By choice of a suitable infusion rate it is possible to use the new method to study metabolic rates at substrate concentrations which it was not previously feasible to use.The major advantage of perfusing the isolated heart with a simple saline medium lies in the control which this affords over the substrates and hormones presented to the heart. Pre-perfusion for 10 to 15 min is commonly used to allow the endogenous hormones and substrates of a freshly excised heart to equilibrate with the perfusate. Thereafter the heart is presumed to be metabolically stable. However, and Fisher and Gilbert [1970b] showed that 30 min were needed in order that the permeability of the isolated rat heart to D-xylose and L-arabinose in the absence of added insulin should decline to a stable minimum. The effect was attributed to the elution of endogenous insulin. It is necessary to be able to measure glucose utilization at short intervals in order to determine whether it declines similarly in the early stages of perfusion. This paper describes a means of doing this in conditions in which the perfusate glucose concentration is almost constant.A new apparatus for cardiac perfusion has been developed which enables a stable metabolic state to be readily established and recognized. A substrate is continuously infused into a constant volume of recirculating perfusate. The system tends to a steady state in which the composition of the perfusate is constant while the heart remains metabolically stable. The properties of the system allow the time-course of utilization of a substrate to be followed continuously. An initial decline in glucose utilization in the absence of added insulin was observed when the perfusate glucose concentration was in excess of 1 mm, but not otherwise. The decline continued for up to 1 hr, suggesting that an insulin-free heart has rarely been studied. When insulin was added initially at high concentration (100 mU/ml.) the glucose utilization was steady throughout the perfusion. We conclude that the decline in the absence of 176