Viability detection is crucial for the management of myocardial infarction (MI). Signal intensity (SI)-based MRI methods may overestimate infarct size in vivo. In contrast to SI, the longitudinal relaxation-rate enhancement (ΔR1) is an intrinsic parameter that is linearly proportional to the concentration of contrast agent (CA). Determining ΔR1 in the presence of an infarct-avid persistent CA (PCA) allows determination of the per-voxel percentage of infarcted tissue. Introduced here is a ΔR 1-based CE-MRI method, termed percent infarct mapping (PIM), for quantifying myocardial viability following delayed PCA accumulation. In a canine MI model (N = B), PIMs were generated using a persistent CA (PCA) and validated using triphenyltetrazolium-chloride (TTC) histochemistry. Voxel-by-voxel R 1 maps of the entire left ventricle (LV) were generated 24 and 48 hr after PCA administration using inversion recovery (IR) with multiple inversion times (TIs). PI values were calculated voxel by voxel. Significant correlations (P < 0.01, R = 0.97) were obtained for PI per slice (PIS) determined using PIM vs. corresponding TTC-based values. Median deviations of PIS with PIM from that with TTC were only 1.01% and -0.53%, at 24 hr and 48 hr. Median deviations from the true infarction fraction (IF) were 1.23% and 0.49% of LV at 24 hr and 48 hr, respectively. No significant difference was found between PIM 24 hr and PIM48 hr. ΔR1-based PIM is an accurate and reproducible method for quantifying myocardial viability distribution, and thus enhances the clinical utility of CE-MRI. © 2006 Wiley-Liss, Inc.