Bird–window collisions at houses have been identified as a significant source of mortality for North American birds, but which types of houses and windows are most problematic remains poorly understood. We assessed how neighborhood type, yard conditions, house attributes, and window type influenced collision rates. Data were collected from citizen scientists across Alberta, Canada, who surveyed their houses daily. In relation to the best-fitting model, the yard model explained 58.1% of the explained deviance, the neighborhood model 45.6%, and the house model 42.6%. The factors that had the largest effect for predicting collision risk included season and whether the house was in a rural or an urban area (rural areas in the fall had a 6.0× higher collision risk than urban areas in the winter), the height of vegetation in the front yard of the house (trees >2 stories high increased collision risk by 3.6× compared to houses with no trees), and the presence of a bird feeder (which increased collision risk by 1.7×). This suggests that multiple factors affect collision rates and that the suitability of a yard as bird habitat is likely a key driver. Given that few homeowners are likely to take an approach that reduces the number of birds in their yards, future focus needs to be given to bird-friendly urban design and developing the most effective window deterrents so that collisions can be reduced and birds enjoyed in urban environments.