Ethnoracial inequality in educational achievement and attainment is a long-standing feature of the American education system. For example, national reading assessments show that White fourth graders scored nearly a standard deviation higher than Black or Hispanic students on average in 2015 (NCES, 2015). The measurement and analysis of differences in achievement among ethnoracial groups are imperfect means to fully understand and appreciate the depth of racial inequality in the U.S. education system. However, unequal outcomes, while lacking nuance, can shine a light on unequal opportunities. Instead of framing educational disparities through a deficit lens, attributing racial/ethnic gaps to the shortcomings of Black or Hispanic students, one can frame educational inequality as the result of opportunity gaps, resulting from disparate social and educational opportunities available to students of different ethnoracial groups (Carter & Welner, 2013). On their own, statistics about racial and ethnic test score gaps provide a blunt description of trends and patterns in these racialized opportunity gaps. Coupled with demographic, institutional, and community-level data, information about differences in achievement can be used to more fully understand ethnoracial inequality in the American education system. Observing and experiencing persistent ethnoracial disparities in school can shape students' beliefs about ethnoracial groups and create barriers to diverse friendships (Allport, Clark, & Pettigrew, 1954; Moody, 2001; Tyson, Darity Jr, & Castellino, 2005). Moreover, achievement outcomes predict long-term educational attainment, employment, and earnings (Chetty, Friedman, & Rockoff, 2014; Currie & Thomas, 2001; Jencks & Phillips, 1998), so these school-age educational inequalities may lead to future ethnoracial economic disparities. Therefore, understanding ethnoracial achievement gaps is imperative for documenting and addressing current and future inequalities. Existing research on achievement inequality documents trends in national- or state-level achievement gaps or notes unequal outcomes at one point in time from survey data. Recent data from the Stanford Education Data Archive (SEDA) allow, for the first time, a comprehensive examination of achievement scores for nearly every public school district in the U.S. These data show that achievement inequalities exist within districts as well as nationally, with White students outperforming their Black and Hispanic peers, on average, in nearly every district in the U.S. (Reardon, Kalogrides, & Shores, 2019). In this article, we contribute to existing literature in three ways. First, we estimate trends over time in test score gaps for each grade in each school district, which we interpret as indicative of district-level change. This is in contrast to measuring changes in achievement equity within cohorts of students over time (Reardon, 2019; Reardon, Weathers, Fahle, Jang, & Kalogrides, 2019). Second, we conceptualize definitions of achievement gap trends that consider how the given district's non-White subgroup fares in relation to White peers within their own school district ("within-district gap trends") as well as to White students nationally ("national gap trends"). Documenting inequalities both within districts and at a macro scale is critical to identifying the pathways through which gaps emerge and the consequences of relative academic achievement for later outcomes, such as the college admissions process. One might reach different conclusions about achievement gap patterns depending on the reference group. Third, we establish methods for detecting meaningful degrees of ethnoracial achievement gap shrinkage over time (again, both within districts and relative to peers nationwide) to understand how much progress toward racial equity is occurring. To do so, we address several methodological and measurement challenges, including scaling of achievement outcomes, estimation approach, data quality thresholds, and precision weighting. Our results provide a comprehensive and rich description of ethnoracial test score gaps in grades 3 through 8 from 2009 to 2016 by addressing the following three descriptive research questions: (1) What is the prevalence of ethnoracial achievement gap shrinkage across the U.S., and how much do gap trends vary in magnitude? (2) Do 'within-district gap trends' covary with 'national gap trends', as defined above? (3) What is the extent of variation in gap shrinkage by grade and subject within a district? First, our findings reinforce that systemic inequality in the educational outcomes of historically-disenfranchised Black and Hispanic student subgroups remains both persistent and pervasive. Among the approximately 9,500 district-and-grade combinations we examined nationwide in 2009, White students had higher average test scores than Black and Hispanic students in the same district in 98-100% of cases. There are virtually no exceptions to this pattern of ethnoracial disparity. This reinforces the fact that the U.S. remains very far from "leveling the playing field" for Black and Hispanic students. Some districts, however, may be making progress toward equity. White-Black within district gaps shrank in 37% and 30% of district-grades in ELA and math, respectively. White-Hispanic within-district ELA gaps shrank in 57% of district-grades, while math gaps shrank in 36% of district-grades. The evidence points to more progress in national gap shrinkage--in more than half of districts, Black and Hispanic students made gains relative to the national mean for White students. That said, even when gaps shrank, the change was modest in size. For example, three-quarters of districts experienced within-district White-Black gap shrinkage of less than 0.10 standard deviations. We also find that less than 10% of districts with any gap shrinkage (in any grade) exhibited gap shrinkage in all six grade levels. The notable variability (across subjects and grades) among gap trends in the same district complicates explanations for why and how districts and their local communities might help reduce academic inequality.