West et al. (1999) present a reasonable argument for the pluralistic view that multiple adaptive mechanisms may simultaneously, and even synergistically, favour sexual reproduction. However, they ignore a fundamental and potentially important distinction between the adaptive role of sex in extant organisms and the evolutionary origin of sex. If one seeks to explain the current utility of sex, then I am inclined to agree with their view that multiple factors contribute to its maintenance and prevalence. On the other hand, I suspect that one selective factor may have been important in the emergence of sexual organisms from an asexual ancestor (although I do not have a preferred candidate for what that single factor was). Even if several factors were involved in the origin of sex, they may have acted sequentially rather than simultaneously, so that each step along the way provided a solution to one problem (e.g. Maynard Smith & SzathmaÂry, 1995). This distinction between multiple factors acting simultaneously to maintain sex, versus a single factor or several sequential factors accounting for its origin, is merely a hunch on my part. It often seems to be the case that evolutionary innovations are initially driven by one primary factor; but after an innovation has been integrated into an organism's way of life, the organism becomes multiply dependent on the continuation of that trait. For example, consider the origin of insect wings during the Devonian period (Kingsolver & Koehl, 1985). Small proto-wings may have been initially bene®cial for thermoregulatory capacity. As the size of these structures increased, owing to change either in relative proportions or overall body size, they subsequently acquired aerodynamic properties that were bene®cial in certain settings. Despite the initially one-dimensional selection to acquire the antecedents of wings, a thorough study of natural selection acting on these structures in extant organisms would probably ®nd evidence to support multiple adaptive roles, including not only thermoregulation and locomotion but also courtship display in certain groups. Thus, selection against loss of a derived trait may often be more complex and multifaceted than was selection to produce that trait originally, as a consequence of the subsequent integration of the trait into the whole being. That this distinction is potentially relevant to the evolution of sex can be seen more clearly by considering the recommendations that derive from the pluralistic viewpoint of West et al. (1999). They suggest focusing future research on a few biological systems to obtain a detailed picture of multiple selective forces and their interactions. They then state that it is `1⁄4 highly important to estimate relevant parameters in sexual species. The form of selection must be different in sexual species than in species which are asexual 1⁄4' This advice is quite sensible if one seeks only to understand the selective factors that are responsible for the maintenance of sexual reproduction in extant organisms. But it can be turned on its head with respect to understanding the evolutionary origin of sex. After all, sexual organisms evolved from asexual organisms that experienced the conditions ± ecological and genetical ± necessary to promote the emergence of sex. The origin of sex can be addressed not only from a historical perspective (using comparative methods and evidence from palaeontology), but it can also be studied by performing experiments with extant asexual organisms to determine whether they ful®l essential preconditions for the emergence of sexuality according to a particular hypothesis. For example, in our own work