The reading achievement scores of over 3,000 students in grades 1 to 3 in 11 inner-city schools were studied to evaluate the effectiveness of the Sullivan Programmed Reading Series. All participating teachers were trained and aided by preservice and inservice workshops, building reading consultants, a speech improvement teacher, an instructional services center, parents, teacher aides, and a consultant from Webster/McGraw-Hill. Base-line data were collected in 1966 using the Gates-MacGinitie Reading Tests for grades 1 and 2 and the Lorge-Thorndike Intelligence Test and the Iowa Test of Basic Skills for grade 3. These same tests were used in the same manner to collect data in 1969 and 1970 for comparing achievement results. A .001 level of confidence was used for all tests of differences. Major findings included (1) reading achievement scores increased significantly for each class entering the project by grade 1; (2) mean achievement scores in vocabulary and reading comprehension (Gates-MacGinitie) for the 1969-70 first grade were at grade level for published national norms; (3) average grade equivalent score growth for second-grade students (1969-70) was more than 1 year in both vocabulary and reading comprehension; and (4) grade scores increased for both the most able readers and the least able readers. Tables and graphs are included. (VJ)