The critical weight (fat) hypothesis has generated interest as a mechanism for the onset and maintenance of normal reproductive cycles in human females. It postulates that menarche is triggered by the attainment of a critical percentage fat and that the maintenance of menstrual cycles requires the persistence of a minimal level; each level is argued to be universal for a particular population or race and, by implication, for the species as a whole. However, an examination of the evidence reveals the hypothesis to have a number of serious methodological and empirical shortcomings which may be grouped under three headings. First, there are no acceptable measures of body fatness, and workers rely instead on estimates from height and weight which have been shown to be intolerably erroneous for individuals. Second, using either these estimates, or more reliable ones using appropriate methods, it may be shown that critical levels do not apply: numerous exceptions involving fat levels above and below the suggested threshold may be shown. Finally, where specific conditions involving reduced fatness and menstrual delay or dysfunction are used to support the hypothesis, it may be shown that other confounding factors are equally plausible. Consequently, based on available evidence, the critical weight (fat) hypothesis cannot be accepted. Hypotheses based on normal maturational processes, especially of the central nervous system, currently provide better explanations.