It is commonly believed that colonization of early-season cotton, Gossypium hirsutum L., by overwintered boll weevils, Anthonomus grandis grandis Boheman, is concentrated on field margins. However, supporting experimental evidence is not available. In 1999 and 2000, we examined colonization patterns of overwintered boll weevils in Central Texas cotton on the bases of adult collections by a pneumatic sampler and hand collections of abscised infested squares. Samples were taken from sites arranged in a grid that extended inward >70 m from the field margin. Adults were collected from shortly after seedling emergence until the flowering stage, and infested squares were collected during the one-third grown square stage. Despite numerical trends, the numbers of adult weevils collected were not significantly different between years or sexes, or among plant phenological stages. Field-to-field variation among collections was considerable and likely prevented detection of differences among these factors. Spatial patterns represented by adult weevil and infested square collections were examined by logistic regressions fitted to the respective probabilities of weevil detection at each designated sample site. Although we observed trends for slightly decreased probability of weevil detection with increased distance from the field margin, these trends were too weak to be demonstrated statistically. Our results indicate the boll weevil does not consistently exhibit a strong edge-oriented colonization pattern, and that management tactics that are predicated on these patterns, such as border sprays, should be used with caution.