Significant cross-sectional associations between mood and weight have been made in women; however, data on associated longitudinal effects and their psychological and behavioral mechanisms are required to inform obesity treatments that mostly have limited success beyond the very short term. Women participating in behavioral obesity treatments were assessed on psychological and behavioral measures, and weight change over 12 months. A treatment focused on physical activity and self-regulation (n = 67) had significantly better improvements than a treatment centered around weight-loss education (n = 64) on measures of mood (overall mood, depression, anxiety), self-regulation, emotional eating, eating behaviors, physical activity, and weight in women with obesity. Incorporating a lagged variable design, 12-month weight loss was significantly predicted (separately) by changes in overall negative mood, depression, and anxiety. When changes in measures of self-regulation, emotional eating, and eating behaviors were sequentially entered as mediators, mood change-weight change relationships were rendered non-significant. Significant mediation paths were: mood change→self-regulation change→weight change, and mood change→self-regulation change→eating behavior change→weight change. They were unaffected by the treatment group. Findings contributed to both theory and obesity intervention architectures via a design sensitive to the dynamic psychological and behavioral changes occurring within weight-loss processes.