Ovarian failure is heterogeneous, both in cause and in phenotype. Women whose menses cease prior to age 40 years and have menopausal FSH levels are traditionally considered by clinicians to have the diagnosis of “premature ovarian failure” (POF). In fact, however, the ovarian function in this disorder is in reality a continuum of impairment in ovarian responsiveness. For this reason the term “primary ovarian insufficiency” (POI) has been suggested to be more accurate. It would be yet more desirable to classify this disorder on the basis of aberrant embryonic differentiation rather than the state of ovarian function. Genes that exert known hormonal effects (FSH, FSHR, LH, LHR, CYP17, CYP19) primarily affect follicle function, as do BMP15, GDF9, and GPR3. These genes mutations have occasionally been found in humans, but none are common. Still other genes expressed during oogenesis appear more likely to be characterized by lack of germ cell formation. These include DNA binding proteins and transcription factors like NOBOX and LHX8, and RNA binding proteins like NANOS. Plausible causative mutations have been identified in a few women (NOBOX, GDF9, LDX8), but even then only 1–2% of cases show a perturbation. Thus, considerable heterogeneity—phenotypic as well as etiologic—exists in ovarian failure, irrespective of whether follicles do or do not develop. Analysis of other genes is necessary, including many likely to show novel mechanisms of action.