The treatment of 138 patients, seventy with tuberculosis of the hip and sixty-eight with tuberculosis of the knee, and the late results have been considered in this paper. These include patients treated during the years 1948 and 1949 when supplies of streptomycin were scanty and experience of direct operations into tuberculous joint lesions slight.

The durations of follow-up of the 138 patients were as follows: one to four years, twenty-four; five to nine years, forty-eight: ten to fourteen years, forty-nine; fifteen years and over, seventeen. Therefore, 40.6 per cent were followed for ten years or more.

The results were better in the group under fourteen years of age than in the group fourteen years old or older. Forty-six of sixty-five (71 per cent) had a full range of flexion on follow-up; ten of seventy-three (14 per cent) of the older age group had a full range of flexion. Seven of sixty-five (11 per cent) of the younger age group had fused joints at follow-up examination; twenty-nine of seventy-three (40 per cent) of the older age group had fused joints. The high rate of fusion of joints in the older age group is partly accounted for by the high proportion of patients in this group who were admitted with advanced destruction in the knee.

The tendency of cartilage to survive even in joints severely affected by tuberculosis was noted; significant return of function might occur even in a joint in which up to half of the cartilage had been destroyed. It is assumed that repair by fibrocartilage or smooth fibrous tissue must have occurred.

It was found that after more than 100 days of chemotherapy, specific or nonspecific changes persisted in the majority of joints, and comment is made that tardy synovial recovery would delay cartilage regeneration. Pannus formation over cartilage was noted, and it is suggested that the persistence of pannus formation also delays joint recovery. The decreasing use of splints and the increasing reliance on early exercise and earlier ambulation are also commented on.