Evolutionary biologists consider all complex design features of organisms to be ultimately the result of natural selection. As such, menopause can always be considered an adaptation. At the same time, it is also recognized that an adaptation is always morphologically, physiologically, and developmentally constrained by an organism's phylogenetic heritage. 18 The question of origin is whether menopause is primarily an adaptation, in the sense that selection directly favored a postreproductive life span in human females, or whether it is an epiphenomenon of selection for efficient early reproduction or physiological constraints preventing prolongation of fertility in the presence of increases in human longevity. The distinction between adaptations, fitness trade-offs, and true phylogenetic constraints is the level of explanation. Appeals to the role of constraints without explanation of why these constraints themselves are not subject to evolutionary change constitute proximate explanations. Although it is useful for heuristic purposes to think in terms of three distinct alternativesadaptation, trade-off, or constraintthe explanations overlap and interact. For instance, a trade-off is part adaptation and part constraint. In addition, the question of menopause cannot be isolated from the question of human longevity. In all three cases, the bottom line is that selection favors early reproduction relative to life span. In this sense, reproductive senescence is always premature. In this paper I critically review the evidence for the adaptation, physiological trade-off and by-product of increased longevity explanations for the origin of menopause.