AB - PURPOSE: To explore psychosocial correlates and predictors of stage transitions in the precaution adoption process model (PAPM) for fruit intake. DesigN: A cohort completed three electronic questionnaires, at baseline (time 0), 35 days later (time 1), and another 32 days later (time 2). SETTING: Secured Internet Web site. SUBJECTS: A cohort of 735 adults was formed from a random sample of an existing Internet panel. The mean age was 37.5 years, 51% were women, and 90% were of Dutch origin. Most respondents (48%) had a medium level of education. MEASURES: Precaution adoption process model stage, risk perception, perception of own fruit intake level, attitude, pros, cons, subjective norms, social support, modeling, self-efficacy, and fruit intake (assessed using a food frequency questionnaire). ANALYSIS: Cross-sectional differences in psychosocial variables and fruit intake across PAPM stages at baseline were analyzed using analysis of variance with Tukey multiple comparisons tests. Predictors of PAPM stage transitions between time 0 and time 1 and between time 1 and time 2 were analyzed using logistic regression analysis. RESULTS: Factors related to attitude and social influences may be important if one is to decide to act, whereas strong self-efficacy may also be required for acting on the decision to act. CONCLUSION: Although the results should be replicated in a larger and more representative sample, the PAPM seems a good framework for studying fruit intake.