As part of a general investigation of the of programmed instruction, high school studencs in the Roanoke Public School system took year-long programmed courses in Algebra I, Plane Geometry, and Algebra 11. The senior author was impressed by the number of spontaneous commencs from chese studencs and cheir teachers pertaining co a subjective feeling of marked improvement in the students' reading ability. The present study investigated the hypothesis that students who take mathematics courses in programmed form will improve their reading ability relative to controls taking the same courses with a conventional texcbook. Ss were those students in Plane Geometry (N = 222), Algebra I (N = 105), Algebra I1 (N = 31) who had taken programmed instructional material plus controls (Ns = 84, 45, 27, respectively). (For a description et the overall design, see Encyclopaedia Britannica Films Rep. No. 2, 1961). At the end of the spring term, all Ss in the experimental and control classes were given the Cooperative English Test, Test C 1: Reading Comprehension, Lower Level, Form Y, except those Ss who were ill or absent from school for some other reason on the day of testing. Data were treated by analysis of covariance, using the STEP, Form A reading scores, obtained from the state-wide testing program of Fall, 1959, as the outside variable. Three reading scores were analyzed: total, level of comprehension, and speed of comprehension. Experimental Ss in geometry were significantly better in speed of reading comprehension than their controls (P < .05). None of the other differences were scatiscically significanc. . Utter ( 1961 ) noted that in some experiments at the Bell Telephone Laboratories, Ss who went through programs increased their reading comprehension. The reading test used ac Bell was the Davis Reading Test (R. F. Utcer, personalcommunication, February 5, 1962). In the study conducted by Ucter himself at Sandia, using the Bell Laboratories program in basic electricity, there was no rise in reading comprehension. Utcer speculates that this is probably due to the fact that many Ss had initially high reading scores. Our positive resulc in geometry, and that from Bell Laboratories, indicate some evidence for an interesting side effect pertaining co reading compre-a hension, which may occur during exposure to programmed instructional material. The failure of our Ss caking the algebra courses and those at Sandia to show significant increases in reading comprehension indicates that such an increase is related to specific attributes and does not invariably accompany the use of programmed instructional material.