The extent to which national performance on an achievement test battery changed over a 4-year period was studied, and the comparative performance of students in school districts taking the test series for the first time was compared with that of students in districts that had used the battery at least once in one or more grades. The seventh edition of the Stanford Achievement Test was used to investigate both issues for 1982 and 1986. After consideration of geographic region, system size, and socioeconomic status, a sample of 116,000 students from kindergarten through grade 12 was selected. Results indicate that national performance did improve in reading, mathematics, and language in the lower grades. Students in lower grades in districts that had used the test series at least once outperformed students in districts that had not used the series before. Use of the 1982 norms to assess current students would overestimate the standing of current students relative to the national population, particularly in mathematics and language. Three tables present the study results. (SLD)