The Robust and Equitable Measures to Inspire Quality Schools, or "REMIQS," project is guided by two questions. The first is, "where are the highest-performing schools in traditional settings that consistently promote improved outcomes for our most underserved students?" The second question is, "what practices, policies, and procedures drive the success of those schools?" For this--the quantitative filtering phase of the project--the research team analyzed comprehensive longitudinal data sets from five states to enhance our understanding of the impact of schools on outcomes for historically marginalized students. We applied filters that home in on the project's target population of learners to identify schools based on their effects on agreed-upon desirable outcomes. This phase of the project was intended to support a subsequent qualitative investigation of the practices, policies, and procedures that foster success in those schools. REMIQS is motivated by the presence of opportunity gaps that continue to persist for students from historically marginalized backgrounds. For the quantitative filtering phase, REMIQS defines historically marginalized students to include students who identify as Black, Latinx, Indigenous or multiracial, students from low-income families, students who qualify for special education services, and English learners. As such, the focus on these student groups is in line with the project's aim to pursue social justice by increasing educational equity for historically marginalized student populations (Anderson et al. 2020). Another motivation for the project was the creation of interstate, or pooled, models to make school-to-school comparisons across states. As a result of variation in the data provided across state longitudinal data (SLD) systems, the study team produced models to support within-state school-to-school comparisons. To conduct the analysis our team collected SLD systems data across five participating states. The five states included in the analysis were Arizona, Kentucky, Massachusetts, Texas, and Virginia. The research team established agreements or memorandums of understanding (MOUs) with state education agencies, higher education coordinating boards, and departments of statistics to acquire data from participating states and locate publicly available data from state websites. The samples requested across states were grade 9 cohorts in the 2010-2011 through the 2013-2014 school years. Student-level data was requested in addition to cohort-level indicators. To conduct the analysis the research team first applied filtering criteria aligned with the main objectives and target population for the project. These criteria were selected to ensure that identified schools were serving historically marginalized student populations, and that they would be more typical of the high schools attended by students with these backgrounds. The criteria were used to identify the target population of students and schools and create the analytic data set within the respective participating states. Next, the team t a series of random effects, or hierarchical, models to estimate school effects on outcomes of interest. These models were used to exploit the structure of the data and account for the magnitude of variation in outcomes observed within schools (Sullivan et al. 1999). The school effects were in turn weighted and combined into composite measures of school quality. These measures were used to identify the highest quality schools as per the outcomes of students from historically marginalized backgrounds. After identifying this subset of high-quality schools, the list was further winnowed to select five (one per state) REMIQS schools for in-depth case studies of their success. Similar schools that performed at and below expectation were also identified via propensity scores. These schools were identified to provide a point of reference to contrast practices between the high achieving schools, and schools that operate within similar school contexts. The intent of this feature of the design is to uncover and contextualize what sets these schools apart from their peers. Results from the quantitative filtering phase showed that schools with larger historically marginalized student populations often sent larger shares of those students to two-year institutions compared to schools with smaller historically marginalized student populations. The trend raises larger questions around differences between school settings, and how those differences can influence low-income and underrepresented minority students' decisions to enroll in colleges that are less selective than colleges that their academic credentials would suggest they are qualified to attend (Bowen et al. 2009). Findings also suggested that schools with smaller historically marginalized student populations tended to perform better on our composite school quality measure. The results will inform the next phase of the investigation and equip qualitative researchers with a better understanding of how schools serve students under different conditions and across settings.