Objective To analyse the independent relations between poverty status and cigarette smoking prevalence and cessation in the United States, 1983-1993.

Design An analysis of eight cross-sectional national surveys.

Setting The United States, 1983-1993.

Participants 236,311 civilian, non-institutionalised adult residents of the United States, aged 18 years and older.

Main Outcome Measures Probability of current cigarette smoking and proportion of former smokers among ever-smokers (quit ratio) in surveyed subjects below the poverty threshold, compared with those at or above the poverty threshold.

Results The odds ratio for current smoking among persons below the poverty threshold ranged from a low of 1.10 in 1985 to a high of 1.45 in 1990, and remained between 1.26 and 1.30 during 1991-1993. The odds ratio for smoking cessation (quit ratio) among persons below the poverty threshold ranged from 0.81 in 1985 to 0.64 in 1991, and remained between 0.73 and 0.66 during 1991-1993. These measures of the relations between poverty status and smoking were derived using multiple logistic regression models, which adjusted for the effects of sex, age, education, race, employment status, marital status, and geographic region.

Conclusions Persons below the poverty threshold continue to be more likely than those at or above the threshold both to be current smokers and not to have quit. Poverty may be an indicator of underparticipation in the changing social norms regarding smoking behaviour in recent years. Individuals below the poverty threshold may need focused efforts to help achieve the Healthy People 2000 objectives for reducing adult smoking prevalence. Further understanding of the relation between poverty and smoking is essential to develop effective programmes for this vulnerable population subgroup.