The establishment of large Cattle Egret (Bubulcus ibis) breeding colonies in upland residential areas of east central Texas has been observed since the early 1960s. To help understand why Cattle Egrets choose residential breeding sites and predict where these might occur in the future, the geographic extent and scale-dependent nature of the phenomenon was investigated. In east Texas and along the Texas Gulf Coast, Cattle Egret colonies were found in flooded forests or on islands. However, in east central Texas, colonies were often located in upland areas. A habitat suitability model was constructed at multiple scales, outlining land use classes thought to influence upland colony site selection: water, forest and development/residential. The model classified 79% of upland colony locations in high or very highly suitable habitat and 7% in low or very low suitable habitat. The distribution of classes was significantly different than expected considering the distribution of land cover suitability classes across the entire study area (p = 0.036). Cattle Egrets likely choose upland, residential sites to breed when suitable wetland habitats are limited. When flooded tree and shrub or island habitats are absent, egrets may choose the edges of development for breeding sites to limit potential disturbance from ground predators.