It has been proposed that the chronic inflammatory process of rheumatic diseases is the direct result of the persistence in tissue of toxic mucopeptide-C polysaccharide complexes from Group A streptococcal cell wails (1-3). In this concept, the tissue damage is produced by the direct toxicity of the mucopeptide moiety. This is supported by the observation that the isolated mucopeptide can induce a severe acute necrotic reaction, and the degree of acute toxicity is inversely related to the amount of residual polysaccharide (2). The polysaccharide, therefore, plays the dual role of protecting the mucopeptide from tissue lysozyme and masking the toxic property. Thus, the cell wall fragments can persist in tissue in a relatively innocuous state, and as the polysaccharide is gradually removed by tissue enzymes the mucopeptide is exposed to produce chronic irritation.One experimental model of this toxicity is a chronic, relapsing nodular lesion of the dermal connective tissue of rabbits which occurs over a period of at least 92 days following a single injection of a preparation of Group A streptococcal cell walls (3, 4). Chronic inflammatory lesions showing a comparable histologic process also have been produced in rabbit knee joints (1, 5), and mouse hearts (6). The joint lesions were produced by the intra-articular injection of sterile cell wall fragments and the active inflammation was demonstrable for at least 6 wk. Granulomatous lesions involving valves and myocardium were produced by intraperitoneal injection of much larger doses (about 10 rag) of the cell wall fragments.Heretofore, we have had only indirect evidence that the chronic lesions produced with streptococcal cell walls in these various models are directly related to persistence of cell wall material. This paper describes the relationship of the injected cell wall antigens to the various phases of the relapsing nodular lesion