PURPOSE: To examine associations between decisional balance, self-efficacy, fruit intake, and stage of change transition from precontemplation and contemplation with cross-sectional and longitudinal methods. DESIGN: A longitudinal cohort study with the use of electronic questionnaires was conducted. Three measurements were analyzed cross-sectionally, and the two intervals between the measurements were analyzed longitudinally. SETTING: A random sample of 1500 individuals from an existing Dutch Internet panel resulted in a cohort of 735 individuals. Of the cohort, 648 (response rate 88%), 592 (response rate 81%), and 570 (response rate 78%) respondents completed questionnaires at the start of the present study (T1), 53 days after T1 (T2), and 106 days after T1 (T3), respectively. SUBJECTS: Mean age was 37.5 years, 51% were women, and 57% ate less than the recommended intake of 250 g of fruit per day. MEASURES: Questionnaires included items measuring stage of change, factors favoring (pros) or working against (cons) behavior change, and self-efficacy. A food frequency questionnaire was used to assess fruit intake. RESULTS: Logistic regression analyses showed that pros, fruit intake, and self-efficacy predicted forward stage transition from precontemplation. Self-efficacy predicted forward stage transition from contemplation. Cons did not predict stage transitions. Results from longitudinal analyses were similar to cross-sectional results, except for self-efficacy: no differences between early stages in self-efficacy were found, whereas self-efficacy predicted these early stage transitions. CONCLUSIONS: Within the limitations posed by the sampling frame, results provided support for the Transtheoretical Model, although determinants might not always be stage specific.