The Fagerström Tolerance Questionnaire (TQ) is often used in both research and treatment contexts to evaluate nicotine tolerance and physiological dependence in cigarette smokers. Recently, however, questions about its validity and its usefulness in comparison to other easily collected measures have been raised. In the present study, 100 male subjects reporting for experimental sessions (Sample I) and 50 male and female subjects entering a smoking cessation clinic program (Sample II) were administered the TQ and determinations of plasma cotinine during ad libitum smoking were made. TQ scores were found to be correlated with cotinine levels in both samples, and several of the individual items proved to have statistically significant discriminatory value. Other schemes for determining degree of dependence were considered and found not to be superior to the TQ. Suggestions for further refining the TQ are reviewed.