The transtheoretical model (TTM) posits that processes of change and the pros and cons of smoking predict progressive movement through the stages of change. This study provides both a cross-sectional replication and a prospective test of this hypothesis. As part of a larger study of worksite cancer prevention (the Working Well Trial), employees of 26 manufacturing worksites completed a baseline and 2 annual follow-up surveys. Of the 63% of employees completing baseline surveys, 27.7% were smokers (N = 1,535), and a cohort of these smokers completed the 2-year follow-up. Cross-sectional results replicated previous studies with virtually all the processes of change and the cons of smoking increasing in linear fashion from precontemplation to preparation (all ps < .00001), and the pros of smoking decreasing (p < .01). However, contrary to the hypothesis, the baseline processes of change and the pros and cons of smoking failed to predict progressive stage movements at either the 1- or the 2-year follow-ups. Possible explanations for these findings and concerns about the conceptual internal consistency of the TTM are discussed.