We have previously shown that the myocardial Gd-DTPA concentration ([Gd-DTPA]t(t)) after a bolus injection of Gd-DTPA can be predicted by the Modified Kety Equation (MKE). If [Gd-DTPA]t(t) can be determined by MRI and the data fit to the MKE, then the distribution volume (lambda) of Gd-DTPA and the myocardial flow (F) times the extraction efficiency (E), i.e., the FE product, can be determined. Therefore F can only be quantified if E is known. We measured the global E in vivo in normal canine myocardium, and measured E and lambda, in vitro, locally in normal, acute ischemic (n = 5; coronary artery occlusion < 4 h), infarcted (n = 4; coronary artery occlusion, 6 days) and reperfused (n = 4; coronary artery occlusion 2 h, and reperfusion 2 h and 6 days) myocardium. Results indicate that E differs with F and with individuals and consequently, F cannot be quantified using the MKE unless the local E is also determined in vivo.