A long postreproductive lifespan is a characteristic feature in the life history of human females, which is not shared with other primates. The ultimate cause of menopause has been the focus of much study and has generated a number of evolutionary explanations, most prominently the mother and grandmother hypotheses. Generally, these theories propose that menopause was an adaptive response to changes that led to the divergence of humans from their ancestors, and are based on observations such as the long-dependency time of human infants, early age of weaning of human neonates, high maternal mortality, the supportive role of grandmothers in childcare, and intergroup female transfers. While ongoing debate continues to refine evolutionary theory, the proximate cause of menopause currently receives less attention. Our knowledge about the mechanisms underlying menopause has been largely confined to ovarian exhaustion, the progressive loss of oocytes beginning before birth and continuing to the age of menopause. Most efforts have to date been focused on fitting curves to the few data available, rather than trying to explain why the dynamics of oocyte depletion follows a particular pattern. A few recent studies have attempted to address this problem by demonstrating that oocyte dynamics is a regulated process under tight physiological control, e.g. that ovaries sense both number and quality of their oocytes. In this review we assess our current knowledge from an evolutionary perspective and emphasize the benefit of combining a mechanistic and life history approach.