Scores from the Stanford Achievement Tests administered to 50,000 students in Palm Beach County, Florida, were studied in order to determine whether the speeded nature of the reading comprehension subtest was related to inconsistencies in the score profiles. Specifically, the probable effect of random guessing was examined. Reading scores were considerably higher or lower than corresponding scores on other subtests, especially in grades 4-7. Item analyses were performed on classroom, school, and district-level test results. Examination of completion rates suggested that deviant reading comprehension performance (as compared to other subtests) was related to unusually high or low completion rates. Further analyses indicated that the students who finished the most reading items were not the highest achievers on the other tests. Apparently the speededness of the reading test encouraged random guessing. Norms were made available for the first 40 items of the reading test, thus reducing it from 60 items to 40. Scores on the 40-item test correlated more positively with the other subtests. Moreover, correlations between completion rate and performance on the other subtests were weak. The shorter test resolved the problems associated with the longer test. (GDC)