It has been postulated that antibody may affect tumour growth directly, or indirectly by modulating the activities of T-cells, K-cells and macrophages. Therefore, these studies were designed to gain some insight into whether the pleomorphism exhibited by immunoglobulin classes and sub-classes might account for the numerous roles that have been postulated for antibody in tumour immunity. The model chosen was the inhibition of syngeneic tumour growth (s.e.) by an infection of C.parvrart (i.p.) in CBA mice, and three serological parameters were measured after administration of C.parvum, tumour or both. These were 1) immunoglobulin binding to tumour target cells in vitro (aom.times designated "anti-tumour antibody", in an operational sense); 2) total immunoglobulln class and sub-class levels; 3) snti-C.parvum antibody titres. WLth the aid of serum from appropriately immunised mice, an isotonic antiglobulin assay was developed to detect immunoglobulin binding to target cells in vitro. It soon became apparent that C.parvum (CN 613U) administration (i.p.) to normal mice resulted in the production of immunoglobulin binding to tumour cells in vitro, accompanied by elevated levels of certain serum immunoglobulins (most markedly and consistently IgG^) and high anti-C.parvum antibody titres. These serological changes were dose and route dependent, although not dependent on an intact thymus, and they occurred with other adjuvants. The immunoglobulin binding to tumour cells in vitro was IgM, and exhibited no specificity for tumour cells. C.parvum (CM 613U) administration (i.p.) to tumour bearing mice also elicited immunoglobulin capable of binding to tumour cells in vitro, accompanied by elevated levels of most immunoglobulin classes and sub-classes and high aati-C.parvum antibody titres. Again the changes were route dependent and could he diminished (or, in the case of immunoglobulin binding to tumour cells in vitro, abolished) by the administration of gold salts. They were also apparent in thymectomised mice and in mice treated with other adjuvants. The immmoglobulin binding to tumour cells in vitro occurred in all classes and sub-classes except IgA, and the 7S immunoglobulin exhibited a degree of specificity for homologous tumour ce2fe, although the 19S did not. Preliminary experiments were also undertaken to see if; 1) the antibody detected by the antiglobulin assay was due to genuine antigenantibody reaction; 2) immunoglobulin binding to tumour cells in vitro could be elicited without the intervention of adjuvant; 3) the serological changes could influence tumour growth; U) tumours of lymphoid origin could affect the immune response to defined antigens. The results are related to findings from other laboratories, and discussed from the standpoint of possible mechinisms of adjuvant action. Suggestions are made for improvement of techniques used and for further work in this area.