Behavioral treatments for diabetes have often been unsuccessful and may benefit from a better understanding of the relative effects of two common treatment foci - decreased weight and increased volume of physical activity - on blood glucose. Overweight and obese adults (N = 59; Mage = 60 years) with hemoglobin A1c (HbA1c) values consistent with diabetes participated in a 6-month community-based behavioral treatment based on tenets of selfefficacy theory and social cognitive theory. The treatment was associated with significantly increased physical activity, reduced body mass index (BMI), and reduced HbA1c levels (ps <.001). Changes in BMI and physical activity accounted for a significant portion of the variance in change in HbA1c, R2 =.13, p =.023. Change in volume of physical activity, β = -.36, p =.007, but not change in BMI, β = -.03, p =.792, significantly contributed to the variance in HbA1c change that was accounted for. There was no effect based on the sex of participants. Discussion focused on how findings might impact the efficacy, efficiency, and application of behavioral treatments for diabetes management. ©2012 Asociación Española de Psicología Conductual. Published by Elsevier España, S.L.