The molecular basis of PNH is known. Somatic mutation of the X-chromosome gene PIGA accounts for deficiency of glycosyl phosphatidylinositol-anchored proteins (GPI-AP) on affected hematopoietic stem cells and their progeny. However, neither mutant PIGA nor the consequent deficiency of GPI-AP provides a direct explanation for the clonal outgrowth of the mutant stem cells. Therefore, PNH differs from malignant myelopathies in which clonal expansion is directly attributable to a specific, monogenetic event (e.g., t(9;22) in CML) that bestows a growth/survival advantage upon the affected cell. Multiple, discrete PIGA mutant clones are present in many patients, suggesting that a selection pressure that favors the PNH phenotype (i.e., GPI-AP deficiency) was applied to the bone marrow. The nature of this putative selection pressure, however, is speculative, as is the basis of clonal expansion. In many patients, the majority of hematopoiesis is derived from PIGA mutant stem cells. Yet clonal expansion is limited (nonmalignant), and the contribution of the mutant clones to hematopoiesis may remain stable for decades. Understanding the basis of clonal selection and expansion will not only delineate further the pathophysiology of PNH but also provide new insights into stem cell biology and suggest novel therapeutic strategies for enhancing marrow function.