The majority of existing research on mobility indicates students do worse in the year of a school move; however, this research has not been successful in isolating estimates of the causal effects and largely does not distinguish between different types of moves. Moreover, the separate grade span literature suggests that structural moves (often excluded from mobility work) mandated by a school’s grade span are also associated with negative performance. In this paper, we obtain credibly causal estimates of the impact of mobility on performance, addressing heterogeneity of moves and endogeneity of moving. We do so using richly detailed longitudinal data for five cohorts of NYC public school students making standard academic progress from grades 1-8. We estimate the impact of attending a new school in a given year controlling for observable student characteristics, including a student fixed effect, and using the grade span of the student’s first grade school as an instrument for mobility. We then examine the impact of both structural and non-structural moves, include interactions with prior move history, control for school quality, and finally explore the long-term cumulative impact of mobility on 8^th grade performance. Taken together, we find that: moving leads to lower performance on standardized ELA and math exams; there are persistent negative effects of structural moves “built into” the school system; non-structural moves can be positive; and the effect of mobility depends on prior move history.