Regression discontinuity (RD) is a popular tool for the analysis of economic policies or treatment interventions. This paper extends the classic static RD model to a dynamic framework, where observations are eligible for repeated RD experiments and, therefore, treatments. Such dynamics often complicate the identification and estimation of longerterm average treatment effects. Previous empirical research with such designs typically ignored the dynamics in the model or adopted restrictive identifying assumptions. This paper studies identification strategies under various sets of weaker identifying assumptions and proposes associated estimation and inference methods. The proposed methods are applied to revisit the effect of Californian local school bonds in the seminal study of Cellini et al. (2010).