Recent evidence points to the early elementary grades as a pivotal point for the development of science learning trajectories and achievement gaps. Using data from the Early Childhood Longitudinal Study, this study estimates the degree to which time spent on science and the breadth of science topics/skills covered predict science achievement in the earliest grades of elementary school. Using regression along with school fixed effects and student fixed effects models, we find suggestive evidence in some models (student fixed effects and regression with observable controls) that time on science instruction is related to science achievement but little evidence that the number of science topics/skills covered are related to greater science achievement. These results are generally consistent across student subgroups. We discuss the implications for early science policy and practice.