BackgroundSmall invertebrate animals, such as nematodes and fruit flies, are increasingly being used to test candidate drugs both for specific therapeutic purposes and for long-term health effects. Some of the protocols used in these experiments feature such experimental design features as lifelong virginity and very low densities. By contrast, the ability of both fruit flies and nematodes to resist stress is frequently correlated with their longevity and other functional measures, suggesting that low-stress assays are not necessarily the only useful protocol for testing the long-term effects of drugs.Methodology/Principal FindingsHere we report an alternative protocol for fruit fly drug-testing that maximizes reproductive opportunities and other types of interaction, with moderately high population densities. We validate this protocol using two types of experimental tests: 1. We show that this protocol detects previously well-established genetic differences between outbred fruit fly populations. 2. We show that this protocol is able to distinguish among the long-term effects of similar types of drugs within two broad categories, stimulants and tranquilizers.ConclusionsLarge-scale fly drug testing can be conducted using mixed-sex high-density cage assays. We find that the commonly-used stimulants caffeine and theobromine differ dramatically in their chronic functional effects, theobromine being more benign. Likewise, we find that two generic pharmaceutical tranquilizers, lithium carbonate and valproic acid, differ dramatically in their chronic effects, lithium being more benign. However, these findings do not necessarily apply to human subjects, and we thus do not recommend the use of any one substance over any other.