To address the presence or absence of school librarians in Kansas public schools, a study using analysis of covariance (ANCOVA) was designed to investigate staffing levels for library media specialists (LMSs), the label used for school librarians in licensed-personnel data in Kansas, and student achievement at the school level. Five subject areas (reading, mathematics, science, history/government, and writing) were examined over a four-year period: 2006-2009. The study examined approximately 2.5 million individual assessment results from 1,389 schools. Researchers found that where schools maintained higher and more stable LMS staffing levels, the Annual Yearly Progress (AYP) data revealed higher proficiency rates. Proficiency differences between no-LMS and full-time LMS conditions were small to moderate but critical with respect to meeting AYP targets. Effect sizes were consistent across grade spans and subject areas, and also consistent with those found in other states' impact studies. The researchers recommend future studies including addressing issues of causality with stratified random samples of students using propensity-score match techniques based on logistic regression and creating indices of contribution by weighting the ANCOVA-based proficiency differences. (Contains 6 tables and 4 figures.)