The majority of human tumor cells have highly aberrant karyotypes, typically ascribed to errors during tumor cell division, potentially linked to a failure of DNA repair, or telomeric insufficiency. Here we discuss another option, that of cell fusion which can lead to the re-assortment of chromosomes during post-fusion mitosis. The observation of hyperdiploid cells has a long history in cancer genetics, but the concept of cell fusion has been difficult to test in practice. Here, we examine the role of cell fusion during normal development, and relate that to potential cellular fusion partners for primary tumor cells. In particular, we describe the potential for stromal partner fusion during metastatic mobilization. The evidence for genetic and cytoplasmic diversity in heterotypic fusion partners is described, together with the new tools available to help the evaluation of this process as a tumor driver.