I have previously assumed that stem tumour cells are 'para-embryonal cells' (PECs) poor or missing in major histocompatibility complex (MHC) antigens. PECs might induce adjoining differentiated hyperplastic cells to also express tumoral phenotype and properties, thus transforming them into 'differentiated para-embryonal cells' (DPECs), MHC-endowed. In such a way, PECs, MHC-lacking, would be automatically surrounded by DPECs, MHC-endowed: this tumour organization was experimentally found by Cordon-Cardo et al in a variety of cancers. Now, I suggest that such a tumour histology might preferentially induce an anti-DPEC T cell immune response which, sparing PECs, might release increasing amounts of DPEC antigens in the peritumour site. DPEC antigens might increase synthesis of specific antibodies and subsequent immunocomplex formation at the peritumour site. Here, abundant immunocomplexes might react through their Fc pieces with CD16 receptors of antibody-dependent cell cytotoxicity (ADCC)-endowed immune cells (natural killer (NK) cells, macrophages, polymorphonuclear cells). These cells would thus be stimulated to secrete their lytic factors before and without their coming into contact with target tumour cells. On the other hand, abundant immunocomplexes at the peritumour site might massively activate the complement system, thus generating large amounts of C3b. C3b might react with CD11b receptors of NK cells, stimulating them to also secrete their lytic factors in an ectopic way at the peritumour site, thus impairing NK cytotoxicity. In such a way, in the absence of ADCC and NK cytotoxicity, a tumour cell enhancement might easily occur. In the light of these ideas, a strategy for antitumour immunotherapy and vaccine is then proposed.