Introduction The earliest recorded use of pneumoarthrography of the knee was made by Werndorff and Robinson (1) in 1904, and although numerous studies of the method have been made since that time, it has not achieved widespread popularity as a diagnostic procedure, for several reasons: In the first place, the method requires extensive experience so that a good technic may be developed and film studies can be accurately interpreted. Most reported series of cases in this country have been small, and the accumulated experience of any one man has been limited. In the second place, there has been some misunderstanding regarding the inherent danger of the method. Kleinberg (2) has reported the occurrence of air embolus, but in that instance adequate precautions were not taken to make certain that the needle was not in a blood vessel. Actually, this danger is inherent in any subcutaneous or intramuscular injection if such precautions are neglected. In over 800 pneumoarthrographic studies we have had not a single complication; nor were any complications encountered in Oberholzer's series of 1,200 cases (3). Hauch (4) has cited a personal communication from Samuel in which infection was mentioned as a complication. This danger is certainly minimal if proper surgical asepsis is employed. We have had no case of sepsis in our experience. If pneumoarthrography is attempted in the presence of joint effusion, the effusion may be temporarily increased. Since we have found, however, that the occurrence of fluid diminishes the efficacy of the method, we refrain from employing it in the presence of joint effusions so far as feasible. Finally, the value of the method has not been accurately assayed in many large series of cases. There are numerous reports in the German literature, but, as pointed out above, most of the reported series from this country are small. Kleinberg (5), in 1921, reported a small series but was not overly successful at that time. He recommended the method for detection of loose bodies and hypertrophied synovial tissue, but not for detection of loose or injured semilunar cartilages. Bernstein and Arens (6) injected carbon dioxide into the knee joint and likewise showed the usefulness of the method for the diagnosis of synovitis but not for semilunar cartilage abnormalities. Rechtman (7) emphasized the value of pneumoarthrography as affording an opportunity for a definite pathologic diagnosis instead of the somewhat vague “internal derangement.” Operative procedures could thus be accurately planned instead of being merely exploratory. Bircher (8) reported 250 cases in which 3 to 5 c.c. of abrodil (20 per cent) were combined with oxygen for contrast, giving good diagnostic results. The use of lipiodol in the knee is contraindicated by the study of Burman, Tunick, and Pomeranz (9).