Background: Much has been made of whether the evidence supporting mindfulness-based interventions (MBIs) is consistent or high-quality enough to merit their wide implementation, and numerous meta-analyses, reviews, and meta-analyses of meta-analyses have been conducted in pursuit of this question (e.g., Bockmann & Yu, 2023; Goldberg et al., 2017; 2022; Roeser, Galla, & Baelen, 2022; Segal, Vyas, & Monson, 2021). There are plausible mechanisms of action that may be operating in school-based mindfulness programs (SBMPs) that are different from traditional social and emotional learning (SEL) approaches, that might justify an expectation of a tighter association with academic outcomes. Specifically, whereas "lesson-based" SEL uses an "outside-in" or metacognitive approach (e.g., imagining social scenarios and how one might handle them), SBMPs use shorter but more frequent exercises (e.g., focusing on the present moment using breath and gentle movements) intended to channel increased calm and attention from the "inside-out" into closely followed academic activities, as well as to train attention skills over the long run. Ideally, a comprehensive SEL program would address both "directions" to optimize student outcomes but at present, the potential academic effectiveness of SBMPs has not been adequately studied (Roeser et al., 2022), especially in the youngest grades during which children's identities as learners are being formed. Purpose: The present study investigated the hypothesis (pre-registered) that the academic impacts (math performance, verbal performance, and executive function) of a K-2 enhancement of the Calm Classroom program (CCK2; Luster Learning Institute, 2008) would be stronger than for an active control group (Dunning et al., 2019) employing the district's default SEL program, Second Step (Committee for Children). We also examined the data in an exploratory fashion (not pre-registered) to assess subgroup impacts that may not have appeared in the omnibus tests. We were especially interested in exploring whether students who: started the program at a younger age (kindergarten vs. 1 grade); came from lower SES bands; and started with lower scores on the academic outcome measures, would have higher impacts from the program. The exploratory analyses results will be available by conference time. Setting: Chicago Public Schools identified a pool of schools that were next in the district's plan to roll out school-wide SEL to every school. The schools also had to meet grant criteria by having 70%+ economically disadvantaged students and not being rated at CPS' highest level of school quality. Participants: 16 schools were randomly assigned to CCK2 and 16 schools were assigned to the district's default program, Second Step. (Although Second Step currently has a mindfulness module, none of the schools implemented it during the study years.) Consistent with the segregation in CPS, approximately half of the schools were 95%+ Black; approximately 1/3 were 85%+ Latinx, and the remainder were racially mixed. Two of the Second Step schools dropped out for a final total of 30 participating schools. A total of 1,122 students consented to participate in the study and provided at least pre-test data on the academic outcome measures. Intervention: The study focused on an enhancement of Calm Classroom (CC) for kindergarten through second grade (CCK2). In addition to 3x/day mindfulness exercises read by the teacher, CCK2 also included weekly 30-minute SEL lessons, a tablet app displaying 2-minute nature videos for use in the classrooms' "calm corner" (Moreno, Baker, Varey & Hinze-Pifer, 2018), twice yearly parent engagement meetings, and various ancillary resources (e.g., children's literature, posters displaying diverse children doing mindfulness exercises, comfort objects). CCK2 teachers received approximately 11 hours of training and ongoing support per year. Each school remained in the study for two years. Data collection focused on K and 1st grade in Y1, and 1st and 2nd grade in Y2. Research Design: The study employed a cluster randomized design (using blocking procedure, which will be described in detail in the paper) with active control group. Multiple methods were used to assess and minimize systematic bias, including detailed balanced tests on 16 variables including individual demographics, census tracts data, and standardized testing administered by the district. None of these balance tests showed significant differences across treatment and control at baseline. Data Collection and Analysis: Impacts were assessed using the letter-word recognition and applied problems subscales of the Woodcock-Johnson tests of achievement, and with administrative testing using the Measures of Academic Progress (MAP) tests. Executive function was measured via the Flanker task. Students with complete data were tested three times--at baseline (between October and December), Spring of their first year, and Spring of their second year. To account for the nested nature of the data we employed OLS with standard errors clustered at the school level; results do not meaningfully change when estimated as a HLM. Covariates included baseline score, gender, race, home language, special education, and grade. Results: Main effects for the confirmatory analyses are presented in the table below, in which the displayed coefficients can be interpreted as effect sizes. As shown, verbal/literacy performance demonstrates some significant advantage for the CCK2 group, especially for the Woodcock-Johnson testing; math performance shows a small, non-significant advantage in the direction of the CCK2 group, again, with slightly higher effect sizes for the Woodcock-Johnson measure; and executive function as measured by the Flanker test shows near-zero impacts. Overall, effect sizes are higher for CCK2 impacts after one year of programming than after two years, which was contrary to predictions. Conclusions: CCK2 appears to have a small, unstable advantage over the active control group, especially in terms of verbal performance. We expected the differences in mechanism of action, i.e., EF, would be stronger than they would be in academic performance, however the reverse was true. This could be because our EF measure was not well-suited to the mechanism of mindfulness exercises in that it assessed visual attention rather than auditory attention. The lack of differences in academic performance may help to assuage some stakeholders' concerns that the 3x/daily exercises of CCK2 (vs. the more common SEL format in the Second Step group of once weekly lessons) would be disruptive to instruction effectiveness.