Malaria is a major public health problem in Africa. Traditional methods of controlling malaria no longer provide adequate protection against transmission, and future approaches likely require a combination of technical solutions and behavioral change. We use a cluster randomized controlled trial to study the impacts of an intervention that combines house screening with a behavioral intervention based on health messaging. While house screening provides modest positive benefits, these benefits can be leveraged if it is combined with health messaging. We provide tentative evidence that the impact of messaging varies with the design of the choice architecture: loss-framed health messages seem to do better than gain-based messages––our data suggest they may have larger and more durable effects on behavior and health outcomes. • We organized a large randomized trial on house screening and health messaging in Ethiopia. • House screening reduces malaria prevalence and the number of malaria sick days. • Complementing house screening with a loss-based health message increases impact. • Loss based health messages changes health-producing behavior of screened households. • A simple cost-benefit analysis suggests that screening combined with loss-based messaging is efficient.