In two independent studies, we explored the usefulness of three self-report measures of tobacco dependence--the Fagerström Tolerance Questionnaire (FTQ), the Fagerström Test for Nicotine Dependence (FTND), and the Heavy Smoking Index (HSI). The FTND is a revised version of the FTQ. The HSI is identical to a two-question subset of the FTND. Study 1 involved 932 participants in a seven-session, five-week, group smoking cessation program, and it looked at the ability of these self-report tests to predict expired air carbon monoxide (i.e., heaviness of smoking) at beginning of treatment and cessation at end of treatment. Study 2 involved 1877 participants in a self-help smoking cessation program, and it looked at the prediction of cessation at 16-month follow-up. All tests made statistically reliable predictions of smoking cessation, but generally accounted for little variance (about 1%). In Study 1, the test scores were associated positively with carbon monoxide levels. The shorter (six vs. eight questions), more reliable FTND is to be preferred to the FTQ; and the HSI (two questions) works as well as the FTND. Evidence is presented that suggests that samples of high-scoring smokers will not be well differentiated from the mid-range to the high-end of the scores.