Abstract Background Higher levels of screen use are linked to lower academic achievement in school-aged children and youth. Few studies have investigated the longitudinal associations between different types of early screen use (0 to 8 years) and later academic achievement. Objectives To examine the association between different types of early screen use and academic achievement in Grades 3 and 6, as measured by the Ontario provincial standardized tests. Design/Methods A longitudinal cohort study was conducted among children in the TARGet Kids! primary care cohort in Ontario, Canada between 2008 and 2023. Participant data were linked to Grades 3 and 6 provincial annual standardized test results in reading, writing, and math from 2012 to 2023. Exposures were parent-reported child total screen time, TV/digital media time, and video gaming time, collected prior to academic achievement test. Academic achievement outcomes for Grades 3 and 6 were categorized as below, at, or above the provincial standard for each subject area. Proportional odds mixed effects model accounting for family-level random effect was used to examine the association between each type of screen use and ordinal achievement in each subject area adjusting for confounders. Models were analyzed for the total sample and stratified by child sex. Results This study included 3,322 Grade 3 children (52% male) and 2,084 Grade 6 children (51% male). Screen use was measured at 5.5 years (SD=2.4) for Grade 3 children, with a mean of 1.6 hr/day, and at 7.5 years (SD=2.9) for Grade 6 children, with a mean of 1.8 hr/day. Males had higher screen use, while females outperformed males in reading and writing in both grades. Each additional hour of total screen time was associated with approximately a 10% decrease in the odds of higher achievement in Grade 3 reading (OR=0.91, 0.86-0.96, p=0.001), Grade 3 math (OR=0.91, 0.86-0.96, p<.001), and Grade 6 math (OR=0.90, 0.84-0.96, p=0.002). Similarly, higher levels of TV/digital media were associated with lower achievement in reading and math in Grade 3 and math in Grade 6. Among females, all screen types, especially video gaming (OR=0.55, 0.35-0.86, p=0.009), were associated with lower achievement in math in Grade 3. Conclusion High levels of parent-reported total screen time and TV/digital media in early childhood were associated with lower reading and math achievement in elementary school. Females with high video game use had lower math achievement in Grade 3. Our findings highlight the need to develop and test early interventions to reduce total screen time and TV/digital media exposure, considering the sex-specific associations, to enhance academic achievement in elementary school.