Students entering high school face many new academic challenges. One of the most important is their ability to read and understand more complex text in literature, mathematics, science, and social studies courses as they navigate through a rigorous high school curriculum. The Regional Educational Laboratory (REL) Northwest conducted a study to examine the effectiveness of a teacher professional development program called Project CRISS, which stands for Creating Independence through Student-owned Strategies. Through Project CRISS, high school teachers learn how to apply research-based learning principles and reading/writing strategies in all major subject or content areas using materials, training, and follow-up support provided by the developer. The pretest and posttest were administered in the fall and spring, respectively, of the second implementation year of the treatment. It was hypothesized that by year 2 of the treatment, teachers would have had familiarity with how to use the Project CRISS instructional methods, given the training and guided coaching they received during the first year of implementation. During the second year, when student impact was assessed, the model specified that teachers would continue to receive some training from the national trainer and technical assistance from the district-certified, in-school local facilitator. There was no statistically significant difference between the treatment group and control group on the mean reading comprehension scores. Because random selection was conducted within a blocking design defined by cohorts, poverty status, and state, a sensitivity analysis was performed to determine whether explicit modeling of the effect size variability across blocks would alter the substantive result. The substantive result remained the same: there was no statistically significant treatment effect. In addition to the main impact analysis, one exploratory research question was addressed: does the impact of Project CRISS on student reading comprehension differ for boys and girls? Concerning the moderating effect of gender, no statistically significant impact difference was found between boys and girls. Appendices include: (1) Statistical power analysis; (2) Teacher questionnaire for treatment schools; (3) Missing data imputation procedures; (4) Project CRISS full implementation model description; and (5) Complete multilevel model results for impact analysis. (Contains 3 figures, 18 tables and 4 endnotes.)