Background. Depression symptoms and diagnoses are associated with failure to quit smoking in most studies, but not all.

Method. A new measure of internalization (i.e. symptoms of depression or anxiety, or poor mood) was created to investigate whether internalization would predict smoking cessation in 549 smokers from three randomized clinical trials with inconsistent findings.

Results. Predicted item locations based on a map of the construct of internalization agreed with empirical locations based on item response theory. Internalization was highly correlated with neuroticism. Logistic regressions showed that internalization improved upon the predictions of other affect-related measures. High baseline internalization decreased abstinence from smoking at end of treatment and 3 months thereafter. History of major depression (single-episode or recurrent) failed to predict abstinence.

Conclusions. The broad, dimensional construct of internalization as conceptualized herein appears to be an important predictor of smoking cessation.