The purpose of our study was to compare the effect on diabetes control of group management with the advice-educational technique traditionally used in managing obese outpatients with poorly controlled non-insulin-dependent diabetes mellitus (NIDDM). Forty-one patients were randomly assigned to these two treatment programs, and 32 patients completed the 6-mo study. Initially, patients were seen for 1-h sessions at 1- and 2-wk intervals and later at 1-mo intervals. Patients were asked to do home blood glucose monitoring, decrease caloric intake, increase exercise, and if they were taking insulin, to adjust the dose to attain approximate euglycemia and to stabilize food and exercise patterns. The combined groups reduced mean ± SD glycohemoglobin from 10.9 ± 3.1 to 9.4 ± 2.4% (P < .05). Internal Health Locus of Control Scale was negatively and significantly correlated with initial and subsequent glycohemoglobin values (the more internal, the lower the glycohemoglobin). At the end of the study the patients in the group management program had significantly lower blood glucose levels than those given advice and education, but no significant differences in glycohemoglobin values or percentage overweight were observed. One patient had a normal initial glycohemoglobin, and only 4 patients had values in the normal range of 4-6.8% at the end of the study. Better management programs need to be developed for treating obese outpatients with NIDDM.