Data-based instructional programs have proliferated in American schools despite limited evidence of their effectiveness in improving educator practice and raising student achievement. We report results from a two-year school-randomized evaluation of the Achievement Network (ANet), a program providing schools with standards-aligned interim assessments and intensive supports for instructional data use. Survey data show that ANet increased teacher satisfaction with the timeliness and clarity of the data they receive and available supports for instructional data-use and caused them to review and use interim assessment data more often. ANet did not, however, affect their confidence in data use or how frequently they differentiated instruction. Student impact estimates show no overall effect on student achievement in English language arts or mathematics. Despite the lack program effects on student achievement, we find that achievement is positively correlated with our survey-based measures of teacher perceptions and practices around instructional data use. Exploratory analyses suggest that the success of ANet in improving teacher practice and student achievement varies with the pre-existing capacity of schools to engage in data-based instruction. Schools rated by program staff as having a high level of readiness to implement the intervention prior to random assignment experienced positive impacts on student achievement, while those rated as a having a low level of readiness experienced negative impacts. The following are appended: (1) School Screener Scoring Rubric; (2) Year 2 School Leader and Teacher Survey Scale Items; and (3) School Leader and Teacher Survey Impact Tables.