Abstract: The aim of this study is to test whether subtypes exist among smokers in contemplation. Data from 194 adult smokers that participated in a randomized controlled trial testing the effectiveness of a computer-tailored smoking cessation program in Dutch general practices were used for secondary analysis. Cluster analysis was conducted based on baseline scores on pros and cons of quitting and self-efficacy to quit. Clusters were cross-sectionally compared for demographic variables and smoking characteristics with analyses of variance (ANOVA) and Chi-square tests. Logistic and multinomial regression analyses were used for longitudinal comparison for smoking behavior and stage of change at 6 months follow-up. Three clusters were identified: Early, Progressing and Disengaged Contemplators. Clusters differed significantly on all clustering variables (P < 0.001). Disengaged smokers were significantly less addicted than Early Contemplators. Cluster membership was not predictive of outcome measures. No subtype was identified representing the Classic Contemplator, scoring high on both pros and cons of quitting and low on self-efficacy, as found in previous studies among US samples. The predictive validity of the clusters found was limited.