Background. Recent research has demonstrated that smokers are at an elevated risk for psychiatric disorders. This study extends the enquiry by examining: (1) the specificity of the psychiatric sequelae of smoking; and (2) the variability in the likelihood of these sequelae by proximity and intensity of smoking.

Method. Data come from the National Comorbidity Survey (NCS), a representative sample of the US population 15–54 years of age. The Smoking Supplement was administered to a representative subset of 4414 respondents. A modified World Health Organization – Composite International Diagnostic Interview was used to measure DSM-III-R disorders. Survival analysis with smoking variables as time-dependent covariates was used to predict the subsequent onset of specific psychiatric disorders.

Results. The estimated effects of daily smoking varied across disorders. In the case of mood disorders, daily smoking predicted subsequent onset, with no variation between current versus past smokers or by smoking intensity. In the case of panic disorder and agoraphobia, current but not past smoking predicted subsequent onset; furthermore, the risk of these disorders in past smokers decreased with increasing time since quitting. In the case of substance use disorders, current but not past smoking predicted subsequent onset, with no variation by time since quitting or smoking intensity.

Conclusions. The data suggest that smoking cessation programmes would not prevent the onset of mood disorder, as ex-smokers do not differ from current smokers in their risk for these disorders. In comparison, daily smoking might be a causal factor in panic disorder and agoraphobia, conditions that might be preventable by smoking cessation. Additionally, current smoking might serve as a marker for targeting interventions to prevent alcohol and drug disorders.