Since its introduction in the 1860s, gypsy moth, Lymantria dispar (L.), has periodically defoliated large swaths of forest in the eastern United States. Prior research has suggested that the greatest costs and losses from these outbreaks accrue in residential areas, but these impacts have not been well quantified. We addressed this lacuna with a case study of Baltimore City. Using two urban tree inventories, we estimated potential costs and losses from a range of gypsy moth outbreak scenarios under different environmental and management conditions. We combined outbreak scenarios with urban forest data to model defoliation and mortality and based the costs and losses on the distribution of tree species in different size classes and land uses throughout Baltimore City. In each outbreak, we estimated the costs of public and private suppression, tree removal and replacement, and human medical treatment, as well as the losses associated with reduced pollution uptake, increased carbon emissions and foregone sequestration. Of the approximately 2.3 M trees in Baltimore City, a majority of the basal area was primary or secondary host for gypsy moth. Under the low outbreak scenario, with federal and state suppression efforts, total costs and losses were $5.540 M, much less than the $63.666 M estimated for the high outbreak scenario, in which the local public and private sectors were responsible for substantially greater tree removal and replacement costs. The framework that we created can be used to estimate the impacts of other non-native pests in urban environments.