Societies of ants, bees, and wasps achieve a highly efficient division of labour by matching the morphology, physiology, life-history, and behaviour of each female to her role in the colony. Females thus occur in distinct castes (e.g. workers and queens). Castes are an example of polyphenism: the caste fate of developing larvae does not depend on their genotype, but is determined by their environment. Therefore, caste differentiation must involve the differential expression of genes that are shared by all castes. Identifying caste-specific patterns of gene expression thus offers a special opportunity to study the generation of adaptive morphological variation by a fundamental biological process, and is the key to profound understanding of the evolution of sociality. Because of the recent development of new methods (especially microarrays) for identifying differentially expressed genes, a burst of research on caste differentiation is predicted. I propose to study caste-specific gene expression in the fire ant, Solenopsis invicta, with a large ( and gt;20,000 clones) spotted cDNA microarray that was recently developed, sequenced, and annotated at the prospective host institute, the group of Prof. L. Keller at the University of Lausanne. S. invicta is the second social insect species (after the honeybee) and the first ant species for which a large microarray has become available. I will study differential gene expression at three levels: in queen vs. worker larvae; in minor vs. major worker larvae; and in the brai n of (equally old) foragers and brood care workers.