Collagen arthritis has been passively transferred with a serum concentrate from immunized donors to immunologically naive recipients as well as cyclosporin-treated, type II collagen-tolerant rats. These findings point to an important role for anticollagen antibody and appear to rule out a role for cellular immunity to type II collagen in the initiation of this disease. The passively transferred arthritis was a transient lesion in the majority of naive recipients and in the cyclosporin-treated, type II collagen-tolerant rats as well when a serum concentrate was transferred after the cessation of cyclosporin treatment. When cyclosporin-treated, type II collagen-tolerant rats received transfer concentrate while cyclosporin was administered continuously, arthritis was significantly enhanced, and lasted as long as cyclosporin was administered and in the majority of rats up to 2 weeks after the cessation of cyclosporin treatment. These results, together with a rapid clearance of anticollagen antibody from the serum, suggest that anticollagen antibody is not the sole regulatory factor and that a cellular suppressor system, sensitive to cyclosporin, might participate in the regulation of this disease process.