This panel study investigated the effect of potential predictors on per capita domestic water consumption in the wet and dry seasons. A total of 242 urban households with children under age five were drawn from two urban communities, Abuakwa and Nkawie, in the Atwima Nwabiagya District, Ghana. Data were collected from mothers using interviewer-administered questionnaires and analyzed using correlation and stepwise multiple regression. Mean per capita daily water consumption was estimated at 38.97 and 20.70 liters in the wet (n = 140) and dry seasons (n = 235), respectively. The volume of the primary water storage vessel, number of water storage containers, and household size were the most significant predictors in the wet season, constituting 16% of the variation in water consumption. Duration of water storage, household size, number of water service hours, and volume of the primary water storage vessel emerged as the most significant predictors in the dry season, constituting 40% of the variation in water consumption. Further research that considers a wider range of socio-demographic factors, such as the gender of the household head, culture, religion and water use characteristics of each member of the household are recommended.