Wrist arthrography is performed primarily to evaluate soft tissue abnormalities of the wrist including interosseous ligament tears and triangular fibrocartilage complex (TFCC) perforation. However, the reliability of arthrogram findings is still uncertain. Between January 1991 and August 1992, bilateral arthrographies of the wrist were performed on 32 patients with unilateral chronic wrist pain. Of these, 15 had bilateral positive findings; nine, bilateral negative, four were positive only on the symptomatic wrist and another four were positive only on the asymptomatic wrist. The overall positive rate for the symptomatic and asymptomatic wrists was the same (59 per cent or 19/32). There were 13 patients below the age of 30, of whom, four (31 per cent) had positive findings on the asymptomatic wrist. More than half of those patients with positive findings in examination had multiple lesions on either the symptomatic or asymptomatic wrist. Seventeen patients complained of ulnar-sided pain with ten having positive findings on the symptomatic wrist; of these, nine were found to have ulnar-sided pain with ten having positive findings on the symptomatic wrist; of these, nine were found to have ulnar-sided lesions. Howevers, nine of the 17 patients also had positive findings on the asymptomatic wrist, and all had ulnar-side lesions. No definite correlation between radial-sided pain and radial-sided lesion was found. The sensitivity and specificity of the arthrogram was 59 per cent and 41 per cent respectively; the odds ratio was equal to 1. This study demonstrated that wrist arthrography has a relatively low sensitivity and specificity, even in young patients.