Using hierarchical linear modeling, student standardized test scores are analyzed to determine the impact of mentoring first- and second-year teachers on their students' achievement. The contrasting group used for comparison consists of experienced teachers in matched schools, grade level, and content area. The study contains data from 300 teachers in grades 4-10 (196 treatment teachers and 104 in the contrasting group) serving over 6900 students in language arts, mathematics, and science from around the state of Alaska. The dataset is split into the three content areas that were tested, and students with only one teacher per content area are included in the study. Teacher, district, school, and student demographic information are taken into account. Results show that although mentoring new teachers did not bring the students' standardized scores of new teachers up to the same level as students in veteran classes, they are much closer than expected based on past research (statistically significant but very small effect sizes) for Reading, Writing, and Science. In the case of Mathematics, students in classrooms of mentored first- and second-year teachers perform the same as those in classrooms of veteran teachers. Thus, mentoring conducted through the Alaska Statewide Mentor Project shows promising results to start closing the achievement gap typically seen between the students of new and veteran teachers. (Contains 2 tables and 5 footnotes.)