Since July 1970, Arizona has required that a standardized reading achievement test be administered annually to all third-grade pupils. Prior to test administration, data have been gathered on student traits, teacher background and teaching techniques, and characteristics of the school and the school district. Regression analyses were used to examine the relationship between each of the independent variables and reading achievement. During the testing period, the four variables that have had the highest correlation with total reading have been general ability to do school work, socioeconomic status, ethnic information, and language spoken at home. The four variables with the lowest correlation with total reading have been class size, basic phonics program, teacher self-improvement method, and institution where teacher received bachelor's degree. An extensive table listing the independent variables for each year of testing is included. (AA)