Parasite-induced modifications of host behaviour are known from a wide range of host-parasite associations. In many cases, these behavioural changes are thought to be adaptive and benefit the parasite by increasing its probability of successful transmission. However, in many cases, energy spent on host manipulation will not be available for other functions, such as growth. These trade-offs suggest that in the absence of other constraints, natural selection will optimize, and not maximize, the influence of parasites on host behaviour. This argument is developed and expanded into theoretical considerations of the evolution of host behaviour manipulation by parasites. Among populations of the same parasite species or among closely-related species, the optimal investment into manipulation, or optimal manipulative effort (ME*), of individual parasites is predicted to increase as (1) typical infrapopulation size decreases, (2) prevalence increases, (3) the longevity of the infected host, or of the parasite in its host, decreases, (4) passive transmission rates decrease, and (5) parasite fecundity decreases. This evolutionary analysis indicates that ecological and life history variables may have played an important role in the evolution of manipulation of host behaviour by parasites.