The presence and localization of neoantigens induced in cultured cells, infected or transformed with avian tumor viruses (ATV), were studied ultrastructurally on carbon platinum replicas of cell surfaces. The use of antibody, labeled with hemocyanin molecules, provided sensitive detection and analysis of cell surface antigen distribution. The subgroup-specific antigens of the viral envelope were found in considerable amount in the plasma membranes of ATV-infected chick embryo fibroblasts. The distribution of these antigens over the cell surface, evaluated on cells which were prefixed with glutaraldehyde, was found to be diffuse with a greater density on the cell processes in some cells. Reaction of antibody to viral envelope antigens with living ATV-infected cells resulted in a number of patterns of redistribution of membrane antigen-antibody complexes (AAC). Redistribution occurred in symmetrical or asymmetrical modes. The former consisted of randomly oriented aggregates (patches) of AAC over the cell surface. The latter included: (a) linear accumulation of AAC at cell margins; and (b) condensation of compexes into one or more centers of coalescence. These observations could be made on chick embryo cells infected (but not transformed) by avian leukosis virus, or on cells oncogenically transformed by avian sarcoma virus. The regions of coalescence were suggestive of the "capping" phenomenon seen in other systems, and their formation was temporally correlated with endocytosis of labeled AAC and the gradual loss of AAC from the surface. The effects of several biologically perturbing substances on the processes of redistribution were investigated in ALV-infected fibroblasts. Sodium azide, puromycin, actinomycin D, and colchicine had no effect on either form of asymmetrical redistribution. Cytochalasin B (CB) and iodoacetic acid (IAA) appeared to have some effect on the marginal redistribution, and to completely prevent the condensation into foci of coalescence (FC). When treated with these compounds, reacted with antibody at low temperature, washed free of unbound antibody, and warmed at 37 degrees C, cells rapidly cleared their surfaces of AAC. This was not accompanied by formation of FC or endocytosis. In some of these cells, a distribution was observed which suggested a possible centrifugal flow of antigenic sites-perhaps an alternate route for disposal of AAC. None of the drugs tested affected symmetrical redistribution. Repeated attempts at detection and topographical analysis of a tumor-specific antigen on the surface of Rous sarcoma virus-transformed chicken and rat cells have provided no evidence for antibody to such an antigen in the serum of immunized animals. Autochthonous, homologous, and heterologous immunizations of chickens and rats did not produce a detectable antibody response to a virus-specific tumor surface antigen. Preliminary results, however, suggest the expression of an individual-specific (unique) tumor antigen on the surface of Rous sarcoma cells.