The purposes of this study were to design an experimental communications skills improvement program for low-achieving middle-grade pupils, to investigate certain assumed pupil personal-social characteristics on which the program was based, and to assess the effectiveness of the program. Two basic features incorporated into the new program were (1) a nongraded, supplementary group of 10 to 12 pupils which met 2 hours daily and which were taught by an auxiliary teacher employing group processes and individualized instruction, and (2) integrated teaching of listening, speaking, reading, and writing together with utilization of programed readers and other media and methods specifically selected to contribute to positive conditions of learning. Eight classes from grades 4, 5, and 6 received instruction in eight elementary schools in St. Louis County, Missouri, for 15 weeks. Tests given prior to the program indicated that the experimental students differed from "average" students in personality, classroom peer acceptance, and positive classroom behavior. The differences in pretest scores at the .05 level, favoring the control group, disappeared in post test scores. The control group gained only in reading comprehension, but the experimental group gained in punctuation, total language, spelling, and capitalization. (Author/LH)