Amino acid taste has remained largely unexplored in flies. Although there is evidence that flies exhibit enhanced behavioral sensitivity to amino acids when deprived of dietary amino acids, the cellular and molecular basis for amino acid taste is not understood. Our objective was to systematically study amino acid taste and investigate mechanisms underlying their detection.Yeast is the chief source of dietary proteins and free amino acids. We found that amino acids serve as the key cues for yeast feeding preference. Flies showed a strong feeding preference for yeast extract, which was lost when amino acids were depleted from it. Individual L-amino acids were able to evoke feeding preference to varying extents in binary choice assays with sucrose, which was not observed for D-amino acids. As found previously for yeast, feeding preference for amino acids showed sexual dimorphism and dependence on mating state – mated females showed higher preference as compared to males or virgin females.We conducted an RNAi screen targeting ionotropic receptors (Irs) and identified Ir76b, which is widely expressed in taste neurons, to be necessary for yeast preference. Ir76b mutants lose their preference for yeast extract and amino acids, which was restored by expressing Ir76b in the mutant background using the UAS-Gal4 system. Ir76b is highly conserved in insects and we found that its Anopheles gambiae ortholog (AgIr76b) could rescue behavioral defects in yeast and amino acid preference in Drosophila Ir76b mutants, suggesting functional conservation of this receptor across evolution.By expressing GCaMP3 and imaging amino acid-evoked changes in fluorescence, we found that a subset of tarsal Ir76b+ neurons responded to amino acids. Interestingly, we found a functional overlap between amino acid and sucrose-sensing neurons, but not with those that detect water, salt, bitter or acid tastants. This was surprising because Ir76b has been shown to function as an ungated salt channel. We therefore examined the role of other Ir genes as potential co-receptors that gate Ir76b conductance in response to amino acids. One other candidate that emerged from the RNAi screen was Ir20a. We tested the hypothesis that Ir20a blocks salt response of Ir76b by co-expressing the two receptors in sweet taste neurons. We found that Ir76b was sufficient to confer salt response when expressed alone, but the response was greatly reduced in the presence of Ir20a. Conversely, Ir76b alone was not capable of conferring response to a mixture of amino acids, but a response was observed when Ir76b is expressed together with Ir20a.Overall, our results suggest that Ir20a, and possibly other Irs, combine with Ir76b to form amino acid-gated receptors, and offer a simple explanation for mutually exclusive roles of Ir76b in salt and amino acid sensing taste neurons.