BACKGROUND & AIMS: Guidelines advocate use of magnetic resonance imaging (MRI) to estimate concentrations of iron in liver, to identify patients with iron overload, and to guide titration of chelation therapy. However, this recommendation was not based on a systematic synthesis and analysis of the evidence for MRI's diagnostic accuracy.METHODS: We conducted a systematic review and meta-analysis to investigate the diagnostic accuracy of MRI in identifying liver iron overload in patients with hereditary hemochromatosis, hemoglobinopathy, or myelodysplastic syndrome; liver biopsy analysis was used as the reference standard. We searched MEDLINE and EMBASE databases, the Cochrane Library, and gray literature, and computed summary receiver operating curves by fitting hierarchical models. We assessed methodologic quality using the Quality Assessment of Diagnostic Accuracy Studies 2 tool.RESULTS: Our final analysis included 20 studies (819 patients, total). Sensitivity and specificity values varied greatly, ranging from 0.00 to 1.00 and from 0.50 to 1.00, respectively. Because of substantial heterogeneity and variable positivity thresholds, we calculated only summary receiver operating curves (and summary estimate points for studies that used the same MRI sequences). T2 spin echo and T2* gradient-recalled echo MRI sequences accurately identified patients without liver iron overload (liver iron concentration > 7 mg Fe/g dry liver weight) (negative likelihood ratios, 0.10 and 0.05 respectively). However, these MRI sequences are less accurate in establishing a definite diagnosis of liver iron overload (positive likelihood ratio, 8.85 and 4.86, respectively).CONCLUSIONS: Based on a meta-analysis, measurements of liver iron concentration by MRI may be accurate enough to rule out iron overload, but not to definitely identify patients with this condition. Most studies did not use explicit and prespecified MRI thresholds for iron overload, therefore some patients may have been diagnosed inaccurately with this condition. More studies are needed of standardized MRI protocols and to determine the effects of MRI surveillance on the development of chronic liver disease and patient survival.