Background and Context: Social-emotional skills contribute to psychological well-being and achievement (Durlak, et al. 2011), and also forecast long-term outcomes like high school graduation and earnings (e.g., Dweck, Walton, & Cohen, 2011). A recent national survey of 884 principals in PreK-12 shows that 98% of principals believe that all students, regardless of background, would benefit from learning SEL skills in schools (DePaoli, Atwell & Bridgeland, 2017). U.S. schools are giving more attention to SEL because of the important effect of social-emotional skills on attendance, GPA, and long-term educational attainment (Jackson, Porter, Easton, Blanchard & Kiguel, 2020). However, fully understanding the impact of SEL is impossible without understanding the typical growth and development of social-emotional skills. Despite the importance of SEL, no empirical benchmarks exist for typical gains in related skills. Empirical benchmarks prove vital in efforts to monitor progress and evaluate programs (e.g. Hill et al., 2008). They represent estimates of status or growth in the population of interest, and can be used as a yardstick to quantify "typical" scores at a given timepoint or growth over several timepoints. Such benchmarks are critical when outcomes are measured using an arbitrary scale where the units have no intrinsic meaning (e.g., when measuring SEL using a Likert scale). An empirical benchmark illuminates practical significance by comparing the skill observed in a particular sample, study, school, or intervention to norms from a relevant population. While such benchmarks are common for academic achievement (e.g., Hill et al., 2008), few if any exist for SEL, especially for change in skills over time. This dearth of empirical benchmarks for SEL likely occurs because there are so few largescale administrations of surveys of social-emotional skills at regular intervals. CORE, a consortium of California districts serving over 1.5 million students, administers a yearly SEL survey (West, Buckley, Krachman, & Bookman, 2017). The systematic administration of the CORE SEL surveys to students in grades 4 through 12 from 2015-18 offers a unique opportunity to establish empirical benchmarks. In this study, we use these data to benchmark gains between adjacent grades and trajectories across several grades for four SEL skills: growth mindset, self-efficacy, self-management, and social awareness. Purpose and Research Questions: There is one existing study that examined how these skills change over time using the same data (West, et al., 2018). However, existing work falls short of what is needed to establish empirical benchmarks. For example, such studies do not provide means and variances that can be used to compute effect size gains, nor do they examine changes over time using a longitudinal cohort (West et al. [2018] examined change over a number of years, but relied on projections, not actual data). In the present study, we examine two research questions to establish empirical benchmarks for SEL changes over time. Those questions include: 1. What is the expected year-to-year gain in the four SEL skills by grade span? 2. How does each SEL skill change across a four-year developmental timeframe, and across grades 4-12? Sample and Measures: The CORE data include 4th- through 12th-grade students who took the survey at least once during the 2014-15 through 2017-18 school years, resulting in a sample of 361,815 students. The survey included 18 items (4 items for growth mindset, 4 items for self-efficacy, 5 items for self-management, 5 for social awareness). Using these data, we followed individual students over time for up to four years. Specifically, we have six non-intact cohorts (students who were in Grades 4-9 in the first year, 2015, and thus had four years of data) with sizes ranging from [approximately]55,000 to [approximately]65,000 students each. We use these data to examine SEL outcomes longitudinally. Data included represent considerable diversity: [approximately]75% of students are Latinx, [approximately]9% are black, and [approximately]25% are low-income (see Table 1). Research Design and Analysis: We scored the survey item responses using a multigroup longitudinal multidimensional item response theory (IRT) model (Kuhfeld & Soland, 2020). Such scores have a mean of zero and standard deviation (SD) of 1 for the first cohort in the first year (4th grade), and thus all other scores can be interpreted accordingly. To answer Question 1, we produced standardized gain scores by subtracting mean scores in, say, 3rd grade from mean scores in 4th grade and dividing by the pooled SD. Those results can be used as empirical benchmarks for SEL gains by construct and grade span. To answer Question 2, we plotted mean scores by grade and cohort over all four years to examine four-year developmental trends and describe their development from 4th to 12th grade. In addition, we fit latent growth curve models to the data by four-year cohort, and also using an accelerated longitudinal design to estimate growth over the course of grades 4 through 12. Results: Table 2 shows standardized gains between grades as students move through school by skill. This table can be used as the basis for empirical benchmarks. Unlike academic achievement, the SEL skills do not show steady growth across grade levels, but rather fluctuate and sometimes decrease as students progress through school. For example, three skills (self-efficacy, self-management, and social awareness) show declines in the middle school years. Figure 1 shows mean scores by grade and skill to examine longitudinal trends across four years. Both self-efficacy and social awareness show steady declines during middle school grades (social awareness decreases by more than half an SD), then plateau in the high school years. One should note that, as in Table 1, there are great fluctuations in the year-to-year gains for skills like self-management and growth mindset that compose the longitudinal trends, and there is variability by year/cohort. Conclusion: The benchmarks we present can help educators and researchers know what to expect when measuring students' SEL over time. For example, given that some skills show declines in middle, these results suggest that programs targeting outcomes might be doing well if self-efficacy, self-management, and social awareness hold steady, or even show small increases during those grades.