Myocardial perfusion was assessed in nine pigs using ultrafast gradient-echo MRI (.5 T, 15-mT/m gradients) at different levels of myocardial blood flow (range,.005-1.84 ml/min/g), generated either by adenosine infusion or by a mechanical occluder, and measured independently using radiolabeled microspheres, Sixty-four consecutive, EGG-triggered, diastolic, short axis images of the left ventricle were obtained during intravenous bolus injections (n = 30) of .05 mmol/kg of gadopentetate dimeglumine, Relative changes in peak intensity, time to peak intensity, washin slope, correlation coefficient, and cross-correlation coefficient were computed from the time-intensity curves obtained from four regions of interest, namely septal, anterior, lateral, and inferior walls, The values from the inferior wall acted as reference for evaluating relative changes in the other three regions, The cross-correlation coefficient (P < .001, rho = .60) and the peak intensity (P <,001, r =.72) showed the best correlation with myocardial blood now, The washin slope showed a weak positive trend (P <.05), but the low value of r (r =.28) indicated that the use of this parameter to predict now was invalid; the correlation coefficient and time to peak intensity were not correlated (P = ns), In conclusion, this study shows that it is possible to evaluate relative myocardial perfusion after the first pass of a an intravenously injected bolus of gadopentetate dimeglumine, using dynamic MRI on a conventional medium held MRI system. The cross-correlation coefficient and the peak intensity resulted in more efficient parameters to evaluate relative inhomogeneity of regional myocardial perfusion.