Early literacy is a key developmental milestone for succeeding in school, yet far too many children in America, particularly the most socially vulnerable, are not meeting this key milestone. For example, NAEP data from 2013 shows that over 80% of Black and Latinx students are below proficient in reading by the end of fourth grade, as compared to 55% of white students (Annie, 2014). Research shows that students who fall behind in reading by the end of third grade struggle to catch up without significant additional support (Griffin et al., 1998) and are at a much higher risk of dropping out of high school later on (Fiester, 2010). As challenges with literacy are so closely tied to academic performance and success in school in later years, interventions that reduce literacy disparities early on hold promise for generating large and persistent gains in schooling and long-term life outcomes. Numerous literacy interventions, including Reading Recovery, have found that regular individualized tutoring can lead to substantial gains in literacy outcomes among younger students (Shanahan and Barr, 1995). However, the per-pupil cost of these programs is often significant, making it unsustainable for public school districts to provide high-dosage tutoring to children at large-scale. It is in this context that Saga Education worked in partnership with the Chicago Public Schools (CPS) and the University of Chicago Education Lab to implement and pilot-test a lower-cost, high-dosage literacy tutoring model for first graders in two separate CPS charter elementary schools during academic years (AY) 2016-17 and 2017-18. This study aims to evaluate the effectiveness of this literacy tutoring model provided by Saga Education. 1. Research Question: This study focuses on whether the Saga Literacy Model improves first graders' reading skills. For this purpose, we use NWEA MAP scores, a standard reading assessment to measure their progress at the end of first grade. In order to study longer term effects, we also plan to analyze effects on reading scores for participants one- and two-years post intervention. Moreover, we plan to evaluate program effects on our secondary outcomes of interest, including: NWEA reading subtest scores, math test scores, math subtest scores, and attendance. 2. Setting: To test the efficacy of the Saga Literacy Model, the UChicago Education Lab (in partnership with Saga Education and CPS) conducted a two-year RCT in AY2016-17 and AY2017-18. This study took place in 2 pilot schools both of which are charter schools in the LEARN Charter School Network. 3. Participants: The study distributes approximately 300 students first grade students across two study arms, treatment and control across two school years (AY 2016-17 & 2017-18). Table 1 shows the distribution of students across treatment and control for both years as well as the baseline characteristics of the study sample. As can be seen in Table 1, the sample consists of predominantly Black students from low-income backgrounds. 4. Intervention: The Saga Literacy Model personalizes learning by pulling participating students out of regular literacy classes for daily 45-minute tutoring sessions where 2 students work with 1 tutor, similar to Saga's traditional tutoring model (Guryan et al., 2023). The novel idea by Saga Education is to recognize that tutoring is a fundamentally different task from regular classroom teaching and therefore Saga does not require its tutors to have extensive teaching experience. Instead, it leverages the power of dedicated para-professionals, often career switchers or recent graduates. This innovative approach allows Saga to offer high-dosage tutoring sessions throughout the school year at a relatively modest cost. 5. Research Design: This study was structured as a randomized controlled control. We calculate both the intent-to-treat (ITT) and treatment-on-the-treated (TOT) effects on our primary outcome of interest, NWEA reading test scores. 6. Data: The study uses student-level administrative data from study schools and from CPS to measure the effect of the tutoring intervention on our primary schooling outcomes. We use NWEA MAP data to measure the end-of-year impacts on reading test scores, as well as for our secondary outcomes of interest, including: NWEA reading subtest scores, math test scores, math subtest scores. We collect attendance data directly from Saga Education which helps us verify which students participated in programming. To gain a deeper understanding of how the tutoring model was implemented, we also collect observational data on both treatment and control classrooms (167 observations in total). These qualitative observations will complement our quantitative analysis of student test scores. 7. Results: Table 2 presents the results on our primary outcomes for the pooled sample as well as by cohorts for both the ITT and the TOT estimates. Overall, the tutoring program seems to be effective in improving students' reading percentile rank and scores, especially for those who actively participated (TOT model). In particular, effects on reading percentile (4.7 percentile points) and reading scores (1.8 points) are statistically significant for the TOT estimates whereas the reading percentile estimates (4.3 percentile points) are statistically significant for the ITT estimate for the pooled sample. When we compare effects by cohorts, the analysis suggests that participation in the tutoring program led to potential improvements in reading, with a stronger effect observed in cohort 1. Our results are likely an underestimate of true effect sizes since the intervention effectively reduced the class sizes of the control group by approximately one-half. Nevertheless, these impacts are, roughly speaking, comparable to learning gains achieved in the famous Tennessee STAR class-size experiment that reduces early-grade class sizes by a third (Mosteller, 1995). 8. Conclusions: This study reinforces the benefits of high-dosage tutoring, echoing earlier research that demonstrated its positive impact on student math outcomes (Guryan et al., 2023). Here, the high-dosage early literacy model yielded similar effect sizes in reading, suggesting this approach may be effective across content areas and age groups. While this current study is limited given the relatively small sample size, it presents a starting point to increase access to high-dosage tutoring at a lower cost per student. With these promising results, the focus shifts from "what works" to "how can we scale this effective approach?"