Aim In low back pain if serious pathology is suspected diagnostic imaging could be performed. One of the imaging techniques available for this purpose is computed tomography (CT), however, insight in the diagnostic performance of CT is unclear. Method Diagnostic systematic review. Studies assessing the diagnostic accuracy of CT in adult patients suggested having low back pain caused by specific pathology were selected. Two review authors independently selected studies for inclusion, extracted data and assessed risk of bias. Pooled summary estimates of sensitivity and specificity with 95% CI were calculated. Results Seven studies were included, all describing the diagnostic accuracy of CT in identifying lumbar disc herniation. Six studies used surgical findings as the reference standard and were considered sufficiently homogenous to carry out a meta-analysis. The pooled summary estimate of sensitivity was 77.4% and specificity was 73.7%. Conclusions We found no studies evaluating the accuracy of CT for pathologies such as vertebral cancer, infection and fractures and this remains unclear. Our results should be interpreted with some caution. Sensitivity and specificity, regarding the detection of lumbar disc herniation, showed that a substantial part of the patients is still classified as false-negative and false-positive. In future, the diagnostic performance of CT must be assessed in high quality prospective cohort studies with an unselected population of patients with low back pain. © 2011 Springer-Verlag.