Objective Engage grew out of the need for streamlined psychotherapies that can be accurately used by community therapists in late-life depression. Engage was based on the view that dysfunction of reward networks is the principal mechanism mediating depressive symptoms. Accordingly, Engage uses “reward exposure” (exposure to meaningful activities) and assumes that repeated activation of reward networks will normalize these systems. This study examined whether change in a behavioral activation scale, an index of reward system function, predicts change in depressive symptomatology. Methods The participants (N=48) were older adults with major depression treated with 9 weekly sessions of Engage and assessed 27 weeks after treatment. Depression was assessed with the 24-item Hamilton Depression Rating Scale (HAM-D) and behavioral activation with the four subscales of Behavioral Activation for Depression Scale (activation, avoidance/rumination, work impairment, social impairment) at baseline, 6 weeks (mid-treatment), 9 weeks (end of treatment), and 36 weeks. Results Change only in the Activation subscale during successive periods of assessment predicted depression severity (HAM-D) at the end of each period (F1,47=21.05, p<0.0001). An increase of one standard deviation in the Activation score resulted in a 2.04 (95% CI: 1.17–2.92) point decrease in HAM-D. For every one point increase in the Activation score, HAM-D was decreased by 0.22 points (95% CI: 0.12–0.31). Limitations No comparison group. Partial overlap of Activation Subscale with HAM-D, lack of detailed neurocognitive assessment and social support. Conclusion Change in behavioral activation predicts improvement of depressive symptoms and signs in depressed older adults treated with Engage.