This study, using Bandura's Social Cognitive Theory as a framework, tested the effectiveness of a self-regulation program on diet control, exercise and two-hour postprandial blood glucose levels in Thais with gestational diabetes mellitus. Ninety participants were equally divided into an experimental (n = 45) and a control group (n = 45). Both groups were instructed in dietary control and exercise, and received routine care at their respective antenatal clinics. The experimental group members also were trained in self-regulation skills. Instruments for data collection included the: a) Personal Information Questionnaire, and b) Daily Diet Control, Exercise and Two-hour Postprandial Blood Glucose Record Sheet. The Chi-square test revealed the rate at which the experimental group subjects controlled their diet, exercise and two-hour postprandial blood glucose levels, at weeks two, four and six of the program, to be significantly greater than in the control group. Moreover, relative risk analysis (risk ratio) showed the subjects, in the experimental group, could control their diet, exercise and two-hour postprandial blood glucose levels better than those in the control group at weeks two, four and six of the program. These findings illustrate the self-regulation program was effective in helping the experimental group subjects control their diets, exercise and blood glucose levels.