BACKGROUND: We compared the validity of the Cigarette Dependence Scale (CDS-12) and of the Fagerström Test for Nicotine Dependence (FTND) in daily cigarette smokers. METHODS: Internet survey in 2004-2007. Eight days and 6 weeks after answering these two dependence questionnaires, participants indicated their smoking status and answered the Cigarette Withdrawal Scale and the Smoking Self-Efficacy Questionnaire. We used the Mini International Neuropsychiatric Interview (MINI) to assess nicotine dependence as defined in DSM-IV. RESULTS: There were 13,697 participants at baseline, 1113 (8%) after 8 days and 435 (3%) after 6 weeks. CDS-12, but not FTND, predicted smoking cessation after 8 days (odds ratio=1.20 per standard deviation unit, p=0.03) and 6 weeks (odds ratio=1.34, p=0.01). In participants who had quit smoking after 8 days, CDS-12 was a better predictor of craving (beta=0.30, p<0.001), than FTND (beta=0.14, p<0.01). After 8 days, self-efficacy in quitters was predicted by CDS-12 (beta=-0.16, p=0.02), but not by FTND (beta=-0.05, p=0.3). CDS-12 was more strongly associated than FTND with DSM-defined dependence measured by MINI: area under the Receiver Operating Characteristic (ROC) curve=0.72 (95% confidence interval=0.70-0.73). For FTND, the area under ROC=0.64 (0.63-0.66). CONCLUSIONS: CDS-12 performed better than FTND on tests of predictive and construct validity.