The problem with which this investigation is concerned is that of determining whether utilizing self-help groups for remedial mathematics students would improve their course completion rates, achievement, and attitudes toward learning mathematics. The methods of determining the success/failure of self-help groups in this study were the Z-test from inferences concerning two proportions, the t-test from inferences concerning the difference between two independent means, and the t-test from inferences concerning the difference between two dependent means. The participants of the study were chosen from the students enrolled in "daytime" mathematics classes at Tarrant County Junior College - Northeast Campus, Hurst, Texas. The experiment was conducted over two semesters and the data combined for statistical analysis. There were one hundred four students involved in the study. Fifty-two students comprised each of the experimental and control classes. The term self-help group was utilized to describe a small group of two-to-fifteen people who engaged in discussion of responsibility, standards, confession, lay leadership, and action. The students did not study mathematics in self-help group sessions. The group meetings dealt with anxieties, attitudes, and commitment that may be associated with mathematics in general. To investigate the hypotheses of this study, data was collected to calculate the percentage completion rates, the means of the final exams taken by students, and the differences of the Semantic Differential scores given to students in the experimental class at the beginning and the end of the semester. This data was utilized for statistical analysis to determine if the experiment was successful. The report concludes that self-help groups did not significantly improve course completion rates, achievement, or attitudes of students toward learning mathematics. Forty-four per cent of the students that completed the experimental class participated in self-help groups.