We evaluated the diagnostic utility of magnetic resonance imaging (MRI) for predicting anterior cruciate ligament (ACL) tears using both quantitative parameters and nonquantitative imaging findings. MRI examinations were retrospectively evaluated in a group of patients with arthroscopically confirmed complete ACL tear and in a control group with arthroscopically confirmed intact ACL. We evaluated multiple MRI features to compare their sensitivity and specificity for detecting ACL tears. Particular emphasis is put on the evaluation of three different quantitative parameters, including a simplified method for measuring the ACL angle. With a threshold value of 45 degrees the ACL angle reached a sensitivity and specificity of 100% for detecting ACL tears. With a threshold value of 0 degrees the Blumensaat angle had a sensitivity of 90% and a specificity of 98%. Finally, a threshold value of 115 degrees gave the posterior cruciate ligament angle a sensitivity of 70% and a specificity of 82%. Discontinuity was found to be the most useful of the ACL abnormalities. Of the secondary findings anterior tibial displacement was the best predictor of ligamentous injury. However, ACL abnormalities and secondary findings, alone or combined, failed to surpass the diagnostic value of the ACL angle for predicting ACL tears. Quantitative parameters are thus good predictors of ACL tears and may increase the overall sensitivity and specificity of MRI. The ACL angle may be confidently measured in a single MRI section and can be considered to be the most reliable quantitative parameter for detecting ACL tears.