Cytogenetic studies indicate that most tumors are clonal (i.e. unicellular in origin) and have karyotypic alterations. These are not consistent, but non-random abnormalities are being increasingly identified by banding techniques, pointing to the sites on human chromosomes where genes important in neoplastic development are located. It is postulated that tumor progression occurs as a result of genetic lability within the neoplastic clone, leading to emergence of increasingly mutant subpopulations (often recognizable cytogenetically) with more malignant properties. In the context of this hypothesis, acute leukemia, chronic leukemia, and preleukemia can be viewed as differing only in the rate at which an abnormal hemic clone is expanding, with progression to a more aggressive phase (e.g. the "blast crisis" of chronic granulocytic leukemia) reflecting emergence of a new predominant subpopulation as the result of an additional genetic change. These concepts, and the cytogenetic data from which they have been derived, may help our understanding of basic tumor biology, and have some practical applications in the diagnosis of human neoplasms.