The authors present the results obtained in the treatment of rheumatoid knee by arthroscopic synovectomy. A total of 29 knees operated on in 25 patients after a mean follow-up of 38 months based on the Laurin evaluation scale were re-examined. There was significant improvement for pain, synovitis, and overall function (mean values pre- and postoperative 2.07 vs 0.98; 2.14 vs 0.84, 2.37 vs 1.26, respectively); there were no, substantial variations in joint movement (2.47 vs 2.58). By dividing the patients into the 4 Larsen radiographic classes, the worst results were observed in patients in classes 3 and 4. In this series of patients complications of significance were not observed. Thus, we consider arthroscopic synovectomy to be the treatment of choice for rheumatoid knee because of the intrinsic advantages of minimum disease rate, rapid functional recovery, with expectations of results inversely proportional to the stage of progression of the disease.