Communities across the Western United States face the growing challenge of managing water resources in the

face of rapid population growth and climate change. There are two contrasting approaches to understanding

and managing residential water demand in this context. Many scientists and water managers see water use as

a reflection of individual attitudes and decisions where people are assumed to have the agency to act independently of structural constraints. Conversely, other scientists and policymakers focus on the importance of the built environment and the broader social, economic, and policy contexts within which households make water

decisions. Using multilevel models, we compared attitudinal, demographic, and structural drivers of indoor

and outdoor residential water use for a sample of households in Northern Utah. We estimated multilevel

mixed-effect Poisson models with robust standard errors using matched household survey data with metered

residential water use records. Outdoor water use had a substantially greater amount of neighborhood-level variation than indoor water use. Structural factors generally eclipsed individual agency in our analysis.While indoor use was most strongly predicted by household size, tenure status, and length of residence, outdoor water use was most associated with the built environment (lot size and the presence of vegetable gardens and underground sprinklers), socioeconomic status (household income, rental status), and residents' sensitivity to lawn watering norms. Higher water prices were associated with lower water use, with lower-income households being more responsive to prices than higher-income households. Our findings have important implications for water managers and policymakers.