This prospective study, begun in 1984, involves 52 consecutive patients treated arthroscopically for triangular fibrocartilage complex (TFCC) tears. Wrist arthrography showed the tear on the initial radiocarpal injection in 86%, and only on the post-stress films in 14%. Triple-phase bone scan was positive in only 66%. Diagnostic arthroscopy showed linear defects in 34%, a central perforation in 46%, and ulnar or peripheral perforations in 20%. However, 11 patients proved to have no visible TFCC tear (9% false-positive arthrography rate, or 9% false-negative rate for arthroscopy). Of the 41 patients followed for 13-42 months, 88% considered the procedure worthwhile, and 73% had complete relief of pain. The authors conclude that arthroscopic debridement of TFCC tears (often with removal of 2-3 mm of ulnar head) may be of benefit in reducing symptoms without increasing clinical ulnar instability. Wrist arthrography is useful in diagnosing perforations but not useful in evaluating the type, size, or significance of the tear.