The subgroup requirements for accountability in the No Child Left Behind Act (NCLB) were designed to reveal underperformance of disadvantaged groups that could otherwise be hidden in aggregate averages. Both NCLB and its successor, the Every Student Succeeds Act (ESSA), left the choice of minimum subgroup size at the school level (n-size) for accountability purposes to the states. A smaller n-size is more likely to include students from subgroups that are underrepresented at a particular school in considering accountability for that school, but decreases the statistical reliability of the estimate of how students in that subgroup are performing and can raise privacy concerns. Equity-oriented groups that want as many students from disadvantaged groups as possible included in the accountability system, including the Alliance for Excellent Education and the Education Trust, have advocated for states to adopt a minimum n-size of 10, whereas since revoked Obama-era accountability regulations allowed states to choose any n-size up to 30. As the 34 states currently finalizing their ESSA accountability plans for the federal September deadline strive for comprehensive, context-specific strategies to cover more students and schools while maintaining statistical reliability, they should view minimum n-size as just one piece of these strategies. States must consider n-size alongside how they permit schools to combine data over grade levels, school years, and/or groups of students--strategies many states have been using under NCLB waivers and that first-round states have included in their ESSA plans. I use national school-level enrollment data by race/ethnicity to show how many students in different subgroups are covered under different pooling approaches. Pooling data across years and grades will include most students in accountability systems, but for lower enrollment populations, pooling across racial/ethnic groups may provide an opportunity to include students in accountability systems in cases where subgroup size is otherwise too small. Each state should consider the demographic composition of its own districts in making its policy choices, not only about minimum n-size, but also about how districts can combine data to increase the number of students included.