The study's purpose was to examine the measurement domain of confidence items used in physical activity research. We hypothesized that confidence items, including a phrase to hold motivation constant, would differ from standard confidence items. Participants (N = 248 students) completed confidence items, a thought-listing procedure, and a 2-week self-report of physical activity. Results showed that confidence items with motivation held constant loaded exclusively on one factor, but standard confidence items were factor complex with intention. Correlations with physical activity intention and behavior were larger for confidence items than confidence items with motivation held constant. Finally, the thought-listing procedure identified that 3 of the 7 reasons for answering confidence items were outside the intended measurement domain of self-efficacy.