Late synovectomy of the knee joint in rheumatoid arthritis is considered by many authors as a positive influencing factor with regard to the local development of the disease at that particular site. Since this assumption is based on evaluation criteria which are by no means uniform, it has so far not been possible to compare the individual results achieved by different authors. For this reason, we conducted follow-up examinations of 57 knee joints in 44 patients with rheumatoid arthritis in whom the synovectomy had occurred at least five years ago, using the standardised follow-up scheme devised by Gschwent et al. Most of these cases represented a late synovectomy, as was evident from the duration of the disease and from Steinbrocker's stages. An analysis of the results showed that in our patients, pain and swelling were most impressively improved, in accordance with the reports made by other authors. Less impressive was the improvement of mobility following synovectomy. Despite increasing instability, it was usually possible to improve the ability of the patients to walk, especially in cases where this ability had been severely restricted. X-ray films showed in almost one-half of the cases that the objective finding had deteriorated, whereas in the other half of the patients the status was largely unchanged or even improved. We can sum up these results by stating that, taking into consideration the comparable results published in the literature regarding late synovectomy of the knee joint in rheumatoid arthritis, a positive influence exercised by late synovectomy on the local pathology can be assumed.