Psychological research has now clearly demonstrated that there is a significant difference between men and women in their performance on certain spatial tasks. Evidence further suggests that this difference has a neurological basis. This hypothesis is well enough established to have inspired several additional hypotheses concerning the evolutionary origin of the difference, including hypotheses emphasizing male hunting, female foraging, and male reproductive strategy. In this article we examine these hypotheses by placing them against the evidence for the neurological basis for the sex difference and the archaeological evidence for the evolution of spatial thinking in general. Given the probable source of the neurological difference in the timing of fetal testosterone, hypotheses that emphasize selection for female cognitive abilities are handicapped from the start. The hypotheses favoring male hunting and male reproductive strategy stumble when evaluated in light of the timing of the evolution of spatial cognition; archaeological evidence for the proposed selective behaviors and for the spatial abilities in question (e.g., mental rotation) do not correspond in a way that would permit a link between them. We conclude that none of the proposed adaptationist hypotheses fit the evidence as it currently exists, and that the modern sex difference in spatial cognition is almost certainly an evolutionary by-product of selection for optimal rates of fetal development. o 1996 Wiley-Liss, Inc.