We hypothesize that rapid eye movement or REM sleep evolved, in part, to mediate sexual/reproductive behaviors and strategies. Because development of sexual and mating strategies depends crucially on early attachment experiences, we further hypothesize that REM functions to mediate attachment processes early in life. Evidence for these hypotheses comes from (1) the correlation of REM variables with both attachment and sexual/reproductive variables; (2) attachment-related and sex-related hormonal release during REM; (3) selective activation during REM of brain sites implicated in attachment (...) and sexual processes; (4) effects of maternal deprivation on REM; (5) effects of REM deprivation on sexual behaviors; and (6) the REM-associated sexual excitation. To explain why we find associations among REM sleep, attachment, and adult reproductive strategies, we rely on recent extensions of parent-offspring conflict theory. Using data from recent findings on genomic imprinting, Haig (2000) and others suggest that paternally expressed genes are selected to promote growth of the developing fetus/child at the expense of the mother, while maternally expressed genes counter these effects. Because developmental REM facilitates attachment-related outcomes in the child, developmental REM may be regulated by paternally expressed genes. In that case, REM may have evolved to support the “aims” of paternal genes at the expense of maternal genes. (shrink)