AIMS: To examine the relative and combined value of late gadolinium enhancement (LGE) and low-dose dobutamine (LDD) cardiac magnetic resonance (CMR) to predict ‘adverse remodelling’ (AR) following acute myocardial infarction (AMI). METHODS AND RESULTS: Forty-five patients with AMI were recruited. CMR was performed 2-4 days after presentation and at 6 months. Ventricular wall motion and volume were recorded at rest and following dobutamine infusion. Measures of first pass perfusion, persistent microvascular obstruction (PMO), and LGE were obtained following contrast administration. Quantitation was performed using the MEDIS 6.2 software. Regression analysis was employed to determine the univariables and multivariate models most predictive of AR at 6 months. The incremental and relative value of LDD over LGE was investigated. The most predictive univariable was ‘volume of PMO’ (r = 0.51, r2 = 0.26, P < 0.001). The optimal 'combined' multivariate model, utilizing data from all components, was highly predictive of AR (r = 0.82, r2 = 0.67, P < 0.001). The optimal model using parameters only from the LGE component also predicted remodelling (r = 0.65, r2 = 42.0, P = 0.001) but with less accuracy. In contrast, the optimal model using variables from the LDD component alone predicted remodelling with a similar accuracy to the optimal combined model (r = 0.82, r2 = 0.67, P < 0.001). CONCLUSION: A comprehensive CMR examination accurately predicts AR following AMI. LDD is superior to LGE CMR in this respect. These data suggest that LDD not only adds incremental value to LGE in the prediction of remodelling post-AMI but also may be utilized alone with the same predictive power.