Aberrant glycosylation is a hallmark of cancer cells. The blood group precursors T (Thomsen-Friedenreich) and Tn epitopes are shielded in healthy and benign-diseased tissues but uncovered in approx. 90% of carcinomas. T and Tn glycoproteins are specific, autoimmunogenic pancarcinoma antigens. These antigens may also be found in neoplastic blood cells (and on LTV-II infected T lymphocytes). Fundamental chemical and physical aspects of these glycoproteins of primary carcinomas are discussed first. Tn and T epitopes are cell and tissue adhesion molecules, essential in invasion by and metastasis of carcinoma which includes adherent and proliferative phases. These properties are then delineated next, followed by consideration of pathophysiological and clinical aspects of these antigens. Immunohistochemical studies of the extent of Tn antigen expression in primary breast carcinoma demonstrate it highly significant correlation with clinicopathological tumor stages, and hence its value as a reliable prognosticator. On the other hand, there is no significant, prognostically useful association between T antigen expression and clinical disease course. Everyone has "preexisting" anticarcinoma anti-Tn and anti-T antibodies, induced predominantly by the intestinal flora, while cellular immune responses to T and Tn epitopes are evoked only by carcinomas and some lymphomas. Carcinoma dedifferentiation leading to predominance of Tn over T epitopes is described, as is the role of Tn and T epitopes in very early, including preclinical, carcinoma detection. The highest sensitivities in carcinoma detection are for preclinical and the earliest clinical stages. Obviously, preclinical carcinoma detection is of practical importance. T/anti-T tests detected preclinical carcinoma in 77% of 48 patients long (mean 6 years) before their biopsy/X-ray results became positive. There were no false predictions of carcinoma in 38 control persons with benign diseases (observation average 4.8 years). These findings open a novel window for both curative approaches and pathophysiological studies. The autoimmunogenicity of carcinoma T/Tn antigen led us more than two decades ago to begin intradermal vaccination of patients with advanced breast carcinoma of stages IV-IIb, predominately after modified radical mastectomy and sometimes lumpectomy plus axillary dissection always followed by adjuvant radio/chemotherapy. The vaccine consists of human group O red blood cell membrane derived, HLA-free T/Tn antigen containing as adjuvant Ca3(PO4)2 plus a trace of phosphoglycolipid A hyperantigen, i.e., S typhi vaccine (USP), which itself has T and Tn specificities. Our efforts have now for up to 20 years remained successful in a large majority of the 32 patients. All 32 patients survived at least 5 years; 10-year survival was statistically highly significantly improved (5-year survival: P < 1 x 10(-7); 10-year survival: P < 1 x 10(-5)) compared to statistics of the United States National Cancer Institute. Because these vaccinations are successful, their extension to large populations with major types of carcinomas should be considered, and even immunological carcinoma prevention may be contemplated.