his report focuses on the effect of HIV/AIDS on fertility. This effect, in turn, operates at the population level or at the individual level. In one of those paradoxes that demographers, because of their sensitivity to compositional effects, ought to be well equipped to analyze, the population-level and individual-level effects can differ not only in size but also in direction. For example, seropositive status may reduce fertility (i.e. have a negative effect at the individual level) and, moreover, seropositive status may be selective of lower fertility women because of the presence of other STIs. In this case, the relatively higher rates of adult mortality among these lower-fertility seropositive women will, everything else being equal, result in an increase in the level of fertility in the population as a whole. At the individual level, effects on the fertility of infected and uninfected individuals should be considered separately. Among the infected, the many various mechanisms through which HIV/AIDS might affect fertility can be classified as biological or behavioural, whereas among the uninfected only behavioural mechanisms operate.