The impact of a decision by a local program under Chapter 1, the federally funded program of financial assistance to special educational needs of children, to test students fall-to-fall or spring-to-spring was studied. Students enrolled in a Chapter 1 reading program in 1988-89 were tested on a fall-to-spring basis, a spring-to-spring basis, and a fall-to-fall basis; and the results of the three testing points were compared. Thirteen public elementary schools, three parochial elementary schools, and one public high school were involved in Chapter 1 instruction. Spring-to-spring data were available for 319 students (54.3% of the students enrolled in the Chapter 1 program in the district studied). Data for the 1988-89 school year were available for 587 students in grades 2 through 12. The time of testing produced differences in both evaluation and program improvement data. The magnitude of the Normal Curve Equivalent gains varied; different schools were identified for program improvement based on the time of testing. A summer decline in scores was documented for all time intervals. Trends from the spring-to-spring data resembled the no-longer-allowed fall-to-spring data. The requirements of program improvement negatively impacted small and mid-sized schools. Questions were raised about the stress placed on a single normative test score for program improvement. Recommendations were made for changes in the program improvement aspects of Chapter 1. Nine tables present study data, and one figure depicts the reading profile summary for an individual student. (SLD)