Three treatment groups were used to examine the effects on general reading comprehension of teaching and testing four subskills related to the identification of the main idea in prose materials. Subjects were 157 tenth-grade students. Besides a control group, two experimental approaches were used, consisting of subjects who were both taught four main idea subskills and also tested for mastery of them, and subjects who were only tested for mastery of these subskills. The effects of these approaches were measured in terms of subjects' general reading comprehension performances on a posttest and delayed posttest. The results show that a combination of teaching-and-testing main idea subskills does result in a sustained positive effect on general reading comprehension; furthermore, teaching-and-testing masters tend to retain their improved performance while the testing-only masters do not. Eight appendixes provide technical information and sample materials used in the study. (RL)