Information about levels of nicotine dependence in ex-smokers when they smoked, or in current smokers at an earlier date, is useful for clinical and research purposes. To estimate the accuracy of retrospective reports of dependence, 28 individuals who completed either the Fagerström Tolerance Questionnaire (FTQ) or Fagerström Test for Nicotine Dependence (FTND) in smoking cessation trials conducted 5 to 12 years earlier were asked to respond again to the same questions, thinking back to their smoking behavior just prior to their on-study quit attempt. Concordance and Kappa values for the items ranged from 50.0% to 95.0% and 0.00 to 0.92, respectively. The mean difference between the baseline and follow-up total scale scores was 0.05 for the FTQ and 0.38 for the FTND, and the correlation between these assessments was 0.62 for the FTQ (p<0.005) and 0.72 for the FTND (p<0.05). These preliminary results suggest that retrospectively assessed FTQ/FTND scale scores have acceptable reliability.