Teachers require knowledge of the unique skills that each child brings to the classroom in order to effectively target instruction towards students' learning needs. Despite substantial investments in programs aimed at enhancing teacher knowledge of individual students' skills (KISS), we know surprisingly little about how KISS is distributed or how teachers develop KISS, let alone the role that KISS plays in instruction and learning. In this study, I employ nationally-representative data to create KISS measures for kindergarten and first grade teachers. I use these measures to examine the distribution of KISS across schools, within schools, and within classrooms, as well as to investigate potential KISS development mechanisms and instructional uses. To estimate the effect of KISS on student learning, I use a set of student and subject fixed effects models that control for the non-random sorting of students into classrooms, average differences in how well teachers know particular children, and baseline achievement. I find that a standard deviation increase in KISS positively impacts kindergartners' and first graders' achievement by about 0.08-0.09 standard deviations. This result is highly robust to a number of different modeling choices and alternative explanations.