A pilot study systematically collected classroom assessment data in reading and compared that data to the students' test scores on the Degrees of Reading Power (DRP). A total of 61 teachers in 18 New York City public elementary schools gathered the data for 268 students in grades 2 through 6. Teachers identified the books their children could successfully read by listening to the children read aloud, discussing the books, and evaluating student writing about the books. Teachers then looked up the DRP readability values of those books in published guides and compared them to the children's test scores reported in DRP units. Results indicated that the children with the lowest scores could read books that were 10 to 20 DRP units higher on the average than their DRP test scores. Results also indicated that the classroom assessment techniques used by the participating teachers yielded far more information about their students than a standardized multiple choice test such as the DRP. Findings suggest that teachers should collect multiple forms of evidence about student progress in a rigorous, systematic way and that this evidence has consequences both in the classroom and beyond it. This kind of assessment can inform and drive instruction in a positive way while at the same time make schools more accountable. (Appendixes present a letter of commitment, eight tables or charts of data, a sample record of reading performance, a student writing sample, and a sample miscue analysis form.) (RS)