The prevalence of type 2 diabetes and diabetes-related morbidity and mortality is higher among low-income Hispanics when compared to that of Whites. However, little is known about how to effectively promote self-management in this population. The objectives were first to determine the feasibility of conducting a randomized clinical trial of an innovative self-management intervention to improve metabolic control in low-income Spanish-speaking individuals with type 2 diabetes and second to obtain preliminary data of possible intervention effects. Participants for this pilot study were recruited from a community health center, an elder program, and a community-wide database developed by the community health center, in collaboration with other agencies serving the community, by surveying households in the entire community. Participants were randomly assigned to an intervention (n = 15) or a control (n = 10) condition. Assessments were conducted at baseline and at 3 months and 6 months postrandomization. The intervention consisted of 10 group sessions that targeted diabetes knowledge, attitudes, and self-management skills through culturally specific and literacy-sensitive strategies. The intervention used a cognitive-behavioral theoretical framework. Recruitment rates at the community health center, elder program, and community registry were 48%, 69%, and 8%, respectively. Completion rates for baseline, 3-month, and 6-month assessments were 100%, 92%, and 92%, respectively. Each intervention participant attended an average of 7.8 out of 10 sessions, and as a group the participants showed high adherence to intervention activities (93% turned in daily logs, and 80% self-monitored glucose levels at least daily). There was an overall Group x Time interaction (p = .02) indicating group differences in glycosylated hemoglobin over time. The estimated glycosylated hemoglobin decrease at 3 months for the intervention group was -0.8% (95% confidence intervals = -1.1%, -0.5%) compared with the change in the control group (p = .02). At 6 months, the decrease in the intervention group remained significant, -0.85% (95% confidence intervals = -1.2, -0.5), and the decrease was still significantly different from that of the controls (p = .005). There was a trend toward increased physical activity in the intervention group as compared to that of the control group (p = .11) and some evidence (nonsignificant) of an increase in blood glucose self-monitoring in the intervention participants but not the control participants. Adjusting for baseline depressive scores, we observed a significant difference in depressive symptoms between intervention participants and control participants at the 3-month assessment (p = .02). Low-income Spanish-speaking Hispanics are receptive to participate in diabetes-related research. This study shows that the pilot-tested diabetes self-management program is promising and warrants the conduct of a randomized clinical trial.