Abstract The aim of this study was to compare intracoronary (i. c.) administration of Gadophrin-2, a necrosis-avid contrast agent (NACA), and nonspecific agent Gd-DTPA for determination of myocardial viability (MV) in acute myocardial infarction (AMI) with MRI. Reperfused AMI was induced in 12 dogs by transcatheter balloon occlusion of coronary artery. In 6 dogs each, Gd-DTPA at 0.1 mmol/kg or Gadophrin-2 at 0.005 mmol/kg was administered into coronary artery by fast bolus (n = 3) or slow infusion (n = 3). Serial ECG-triggered cardiac MRI of T1-weighted segmented turbo fast low-angle shot (FLASH) sequence was conducted and compared with triphenyltetrazolium chloride (TTC) histochemical staining. The contrast ratio and infarct size were quantified and analysed statistically. No cardiovascular side effects were found with local delivery of both agents. After i. c. administration, Gadophrin-2 induced a strong (CR ≥ 1.78) and persistent ( ≥ 10 h) contrast enhancement of infarcted region. The infarct size defined with Gadophrin-2 was almost identical to that with TTC staining throughout the postcontrast period. With a dose 20 times higher, Gd-DTPA also strongly enhanced infarct-to-normal contrast; however, the enhancement diminished with time, i. e. from early strong to later faint enhancement and eventual loss of contrast. The delineated infarct size was also unstable, i. e. from early overestimation to later underestimation and eventual disappearance of the enhanced infarct. In combination with PTCA procedure, i. c. administration of MRI contrast agents may prove useful for post-procedure verification of diagnosis. The NACA-enhanced MRI may serve as an in vivo surrogate of postmortem histochemical staining for determination of MV. Although applicable in clinical setting, cardiac MRI with nonspecific Gd-DTPA is less reliable and should be performed within less than 1 h after contrast.