Abstract Background Obesity and physical inactivity are responsible for more than 365,000 deaths per year and contribute substantially to rising healthcare costs in the US, making clear the need for effective public health interventions. Calorie labeling on menus has been implemented to guide consumer ordering behaviors, but effects on calories purchased has been minimal. Methods In this project, we tested the effect of physical activity calorie expenditure (PACE) food labels on actual point-of-decision food purchasing behavior as well as physical activity. Using a two-group interrupted time series cohort study design in three worksite cafeterias, one cafeteria was assigned to the intervention condition, and the other two served as controls. Calories from food purchased in the cafeteria were assessed by photographs of meals (accompanied by notes made on-site) using a standardized calorie database and portion size-estimation protocol. Primary outcomes will be average calories purchased and minutes of moderate to vigorous physical activity (MVPA) by individuals in the cohorts. We will compare pre-post changes in study outcomes between study groups using piecewise generalized linear mixed model regressions (segmented regressions) with a single change point in our interrupted time-series study. The results of this project will provide evidence of the effectiveness of worksite cafeteria menu labeling, which could potentially inform policy intervention approaches. Discussion Labels that convey information in a more readily understandable manner may be more effective at motivating behavior change. Strengths of this study include its cohort design and its robust data capture methods using food photographs and accelerometry.