Dam removal is increasingly viewed as a promising restoration strategy. Further research is needed, however, to evaluate ecological responses to dam removal, improve dam removal practices, and develop predictive models of restoration outcomes. Although research has begun to focus on benefits of greater connectivity to fish and other biota, quantitative studies remain scarce. Despite clear benefits, dam removal can also have adverse effects (e.g., mobilization of contaminated sediment, sediment deposition on downsteam biota, spread of exotic species). Researchers can help to identify the circumstances under which these impacts may arise and devise methods that minimize such effects. Two challenges have hampered efforts to predict responses to dam removal. First, some system components (e.g., channel form and riparian trees) may respond very slowly. One approach for estimating these long-term responses is to examine conditions at former dam removal sites. Second, dams vary greatly in size and operation, making it difficult to extrapolate from the results of individual studies. These predictions can be improved by examining how potential responses to dam removal vary with dam characteristics. Finally, dams are often removed to address various socioeconomic concerns (e.g., safety and liability), so more research is needed on these human dimensions of dam removal.