Background: MRI is considered to be the diagnostic tool of choice in diagnosing nerve root compromise among patients presenting with clinical suspicion of lumbo-sacral radiculopathy. There exists controversy among researchers and clinicians regarding the diagnostic utility and accuracy of MRI in detecting nerve root compromise and radiculopathy. This review evaluated 4 primary diagnostic accuracy studies that specifically assessed the accuracy of MRI in detecting nerve root compromise, as established in the current literature. Methods: Eight electronic data bases were searched for relevant articles from inception until January 2014. All primary diagnostic studies which investigated the accuracy of MRI in diagnosing nerve root compromise among patients with low back and referred leg symptoms were screened for inclusion. Qualifying studies were retrieved and independently assessed for methodological quality using the 'Quality Assessment of Diagnostic tests Accuracy Studies' criteria. Results: Four studies qualified for inclusion in this review. The sensitivity of MRI in detecting lumbar nerve root compromise was very low at 0.25 (95 % CI) while the specificity was relatively high at 0.92 (95 % CI). Conclusions: There is lack of sufficient high quality scientific evidence in support or against the use of MRI in diagnosing nerve root compression and radiculopathy. Therefore, clinicians should always correlate the findings of MRI with the patients' medical history and clinical presentation in clinical decision making.