Early MRI and arthroscopic findings were compared to evaluate the efficacy of the former for patients, who upon physical examination, were diagnosed with possihle acute ACL tears. MRI, taken between 1-21 days after injury, had to reveal complete ACL tears and arthroscopic surgery should have taken place 3-12 weeks after the injury for patients of this prospective study. Fifty such cases were initially catagorized according to five MRI patterns. Types I to V were separated by the following ACL tear characteristics: Enlarged and diffusely increased in signal, horizontally oriented, non-visualized, discontinuous and vertically oriented. The cases were then divided into two groups of arthroscopic findings; Group A included cases of complete ACL tearing and group B included those with intact or minimally torn ACLs. Of MRI types I through V, there were 31, 9, 5, 4, and I cases of each respectively. 37 cases resulted in complete ACL tears (21 type I, 6 type II, 5 type III, 4 type IV and 1 type V) and 13 cases were catagorized into group B, 10 of which had pattern type I and 3 of which had pattern type II. As result, types III, IV and V had a 100% Positive Predictive Value (PPV) whereas types I and II together had a 67% PPV vaiue. Division of the cases into the five pattern types in early MRI proved valuble for possible cases of acute ACL tears. MRI pattern types III, IV and V were correlated with definite complete ACL tears. However, caution must be taken for types I and II, the majority of the cases, for which there exists a significant possibility for misdiagnosis and consequently unnecessary arthroscopic surgery.