The proportion of students taking a first algebra course in middle school has doubled over the past generation and there have been calls to make eighth grade algebra universal. We use significant policy shifts in the timing of algebra in two large North Carolina districts to infer the impact of accelerated entry into algebra on student performance in math courses as students progress through high school. We find no evidence of a positive mean impact of acceleration in any specification and significant negative effects on performance in both Algebra I and the traditional followup course, Geometry. Accelerating algebra to middle school appears benign or beneficial for higher-performing students but unambiguously harmful to the lowest performers. We consider whether the effects reflect the reliance on less-qualified teachers and conclude that this mechanism explains only a small fraction of the result.