BACKGROUND: We evaluate whether circumferential strain derived from grid-tagged CMR is a better method for assessing improvement in segmental contractile function after STEMI compared to late gadolinium enhancement (LGE). METHODS: STEMI patients post primary PCI underwent baseline CMR (day 3) and follow-up (day 90). Cine, grid-tagged and LGE images were acquired. Baseline LGE infarct hyperenhancement was categorised as 75 hyperenhancement. The segmental baseline circumferential strain (CS) and circumferential strain rate (CSR) were calculated from grid-tagged images. Segments demonstrating an improvement in wall motion of >/=1 grade compared to baseline were regarded as having improved segmental contractile-function. RESULTS: Forty-five patients (aged 58 +/- 12 years) and 179 infarct segments were analysed. A baseline CS cutoff of -5 had sensitivity of 89 and specificity of 70 for detection of improvement in segmental-contractile-function. On receiver-operating characteristic analysis for predicting improvement in contractile function, AUC for baseline CS (0.82) compared favourably to LGE hyperenhancement (0.68), MVO (0.67) and baseline-CSR (0.74). On comparison of AUCs, baseline CS was superior to LGE hyperenhancement and MVO in predicting improvement in contractile function (P <0.001). On multivariate-analysis, baseline CS was the independent predictor of improvement in segmental contractile function (P <0.001). CONCLUSION: Grid-tagged CMR-derived baseline CS is a superior predictor of improvement in segmental contractile function, providing incremental value when added to LGE hyperenhancement and MVO following STEMI. KEY POINTS: * Baseline CS predicts contractile function recovery better than LGE and MVO following STEMI * Baseline CS predicts contractile function recovery better than baseline CSR following STEMI * Baseline CS provides incremental value to LGE and MVO following STEMI.