The epithelial sheet is a structural unit common to many tissues. Its organization appears to depend on the function of the multi-protein complexes that form adherens junctions. Elegant cell biological experiments have provided support for hypotheses explaining the function of adherens junctions and of their components. These systems, however, lack the ability to test function within an entire organism during development. The realization that the product of the Drosophila segment polarity gene armadillo is related to the vertebrate adhesive junction components plakoglobin and beta-catenin led to the suggestion that armadillo might provide a genetic handle to study adhesive junction structure and function. An examination of the potential function of Armadillo in cell-cell adhesive junctions was initiated using the Drosophila ovary as the model system. We examined the distribution of Armadillo in the Drosophila ovary and demonstrated that this localization often parallels the location of cell-cell adhesive junctions. The consequences of removing armadillo function from the germ-line cells of the ovary were also examined. Germ-line armadillo mutations appear to disrupt processes requiring cell adhesion and integrity of the actin cytoskeleton, consistent with a role for Armadillo in cell-cell adhesive junctions. We have also used armadillo mutations to examine the effects on ovarian development of altering the stereotyped cell arrangements of the ovary. The implications of these results for the role of adhesive junctions during development are discussed.