Objective: Consumption of excess added sugar in the form of sugar-sweetened beverages (SSBs) contributes to a wide range of health concerns in adolescents. Identification of modifiable determinants of SSB consumption based on behavioral theory may inform development of interventions aimed at reducing SSB consumption. The aim of the current study was to test the efficacy of an extended version of the Health Action Process Approach to predict adolescent SSB consumption. Methods: Using a prospective design, adolescent students (N = 450) self-reported their outcome expectancies, perceived behavioral control, intentions, habit, action, maintenance, and recovery self-efficacy, action and coping planning, perceived affordability, and past behavior with respect to SSB consumption. One month later, participants self-reported their SSB consumption. Results: A structural equation model revealed significant direct effects of action self-efficacy, outcome expectancies, and perceived behavioral control on intentions to reduce SSB consumption. Significant direct effects of action self-efficacy on maintenance self-efficacy, and maintenance self-efficacy on recovery self-efficacy, were also identified. There were significant direct effects of intentions and maintenance self-efficacy on action and coping planning. Only intentions and perceived affordability had significant direct effects on SSB consumption. There were also indirect effects of outcome expectancy and perceived behavioral control on SSB consumption mediated by intentions. Inclusion of past behavior attenuated model effects, with past behavior the only remaining predictor of SSB consumption. Conclusions: Findings indicate that adolescent SSB consumption is predicted by intentions and perceived affordability, but effects were extinguished by the inclusion of past behavior. The pervasive effects of past behavior point to the importance of identifying potential mediators of past behavior in future research, and that interventions targeting non-conscious rather than intentional processes may be most effective in reducing SSB consumption.