We developed a dynamic state variable model for studying optimal host-handling strategies in the whitefly parasitoid Encarsia formosa Gahan (Hymenoptera: Aphelinidae). We assumed that (a) the function of host feeding is to gain nutrients that can be matured into eggs, (b) oögenesis is continuous and egg load dependent, (c) parasitoid survival is exponentially distributed and (d) parasitoids encounter hosts randomly, are autogenous and have unlimited access to non-host food sources to obtain energy for maintenance and activity. The most important prediction of the model is that host feeding is maladaptive under field conditions of low host density (0.015 cm2) and short parasitoid life expectancy (maximum reproductive period of 7 d). Nutrients from the immature stage that can be matured into eggs are sufficient to prevent egg limitation. Both host density and parasitoid life expectancy have a positive effect on the optimal host-feeding ratio. Parasitoids that make random decisions gain on average only 35% (0.015 hosts cm2) to 60% (1.5 hosts cm2) of the lifetime reproductive success of parasitoids that make optimal decisions, independent of their life expectancy. Parameters that have a large impact on lifetime reproductive success and therefore drive natural selection are parasitoid life expectancy and the survival probability of deposited eggs (independent of host density), the number of host encounters per day (when host density is low) and the egg maturation rate and number of host types (when host density is high). Explaining the evolution of host-feeding behaviour under field conditions requires field data showing that life expectancy in the field is not as short as we assumed, or may require incorporation of variation in host density. Incorporating variation in walking speed, parasitised host types or egg resorption is not expected to provide an explanation for the evolution of host-feeding behaviour under field conditions.