The aphidophagous midge Aphidoletes aphidimyza (Diptera: Cecidomyiidae) is used in biological control programs against aphids in many crops. Short-term trials with this natural enemy demonstrated that that females prefer to oviposit among aphids colonizing the new growth of plants, leading to differential attack rates for aphid species that differ in their within-plant distributions. Thus, we hypothesized that biological control efficacy could be compromised when more than one aphid species is present. We further hypothesized that control outcomes may be different at different crop stages if aphid species shift their preferred feeding locations. Here, we used greenhouse trials to determine biological control outcomes using A. aphidimyza under multi-prey conditions and at different crop stages. At all plant stages, aphid species had a significant effect on the number of predator eggs laid. More eggs were found on M. persicae versus A. solani-infested plants, since M. persicae consistently colonized plant meristems across plant growth stages. This translated to higher numbers of predatory larvae on M. periscae-infested plants in two out of our three experiments, and more consistent control of this pest (78%–95% control across all stages of plant growth). In contrast, control of A. solani was inconsistent in the presence of M. persicae, with 36%–80% control achieved. An additional experiment demonstrated control of A. solani by A. aphidimyza was significantly greater in the absence of M. persicae than in its presence. Our study illustrates that suitability of a natural enemy for pest control may change over a crop cycle as the position of prey on the plant changes, and that prey preference based on within-plant prey location can negatively influence biological control programs in systems with pest complexes. Careful monitoring of the less-preferred pest and its relative position on the plant is suggested.