Three potential predictors (i.e., biomarkers) of longevity in mammals are proposed: (1) age of pubertal onset, (2) concentrations of gonadal steroids and (3) timing of age-related infertility. Ages of pubertal onset and of declining fertility are hypothesized to be positively correlated with longevity. Concentrations of and androgens and estrogens are proposed to be inversely and positively correlated, respectively, with life span. Evidence upon which these hypotheses are based is reviewed and its limitations are discussed. General strategies for testing these hypotheses are outlined. In addition, several pragmatic as well as theoretical issues facing research on biomarkers of aging are discussed, using specific examples derived from studies of female reproductive aging.