The sensitivity and specificity of bone scintigraphy in the evaluation of meniscus tears was investigated in a prospective study of 44 younger patients referred for arthroscopy with a suspected meniscus tear. We found a sensitivity of 0.74 and a specificity of 0.76 for pure meniscus pathology; thus, the test cannot so far be recommended as a routine in evaluating meniscus pathology. However, the test was highly sensitive if used to divide the patients into 2 groups: a potential surgery-demanding group with meniscus pathology, osteochondritis, degenerative joint disease or rotational instability, and a group with synovitis, hot patella or a healthy knee where surgery (arthroscopy) could be avoided. The accuracy in this division was 0.92. We find the test valuable if used this way.