Small, surface-release dams can impair natural stream connectivity by disrupting the flow of water, sediment, nutrients, and biota. Many dams built during the 19th and 20th centuries have exceeded their functional lifespan, and as a result, dam removal has become an increasingly prevalent stream restoration method. However, many streams have not been consistently monitored before and after dam removal, and there is a paucity of information regarding how dam removals affect stream ecosystems across different dam, stream, and landscape characteristics. Therefore, this study aimed to quantify the effects of dams and dam removals on a critical water quality parameter (stream temperature) across 16 Massachusetts streams with completed (10 sites) or upcoming dam removals. We collected continuous temperature data for 1-3 years prior to removal and up to 5 years following removal in upstream, downstream, and impounded or formerly impounded reaches. Prior to dam removal, most sites experienced downstream warming, although the magnitude and spatial extent of dam impacts on temperature varied across sites. Within one year following dam removal, downstream warming was reduced or eliminated at some, but not all sites. These data suggest that the magnitude and time line of thermal recovery may vary depending on dam (e.g., dam and impoundment size), stream (e.g., upstream temperatures), and watershed characteristics (e.g., impervious/forest cover). An understanding of the factors influencing recovery of stream ecosystems following dam removal may inform our understanding of biotic responses and help set expectations for restoration.