This article traces an emergent tension in an interdisciplinary public health project called Weather Health and Air Pollution (WHAP). The tension centered on two different kinds of data of air pollution: monitored and modeled data. Starting out with monitoring and modeling practices, the different ways in which they enacted air pollution are detailed. This multiplicity was problematic for the WHAP scientists, who were intent on working across disciplines, an initiative driven primarily by the epidemiologists who imbued the project with meaning and value as the protagonists of “health.” To work collaboratively implies a stable, singular, and shared research object, however: one kind of data, one version of air pollution. In detailing two attempts by researchers to address the inadequacies of modeled and monitored data, this article explores the ways in which difference and multiplicity were negotiated and transformed. In doing so, this article suggests that it is the mobility and instability of data that are particularly fruitful for exploring the facilitation and enactment of new realities, while also making explicit the emergent problematics and partialities which inevitably result.