BACKGROUND: Smoking contributes to increased incidence of disease and mortality, as well as to a significant loss in socioeconomic fields. The purpose of this study was to identify the relationship between stress and smoking behavior. METHODS: We analyzed data from 1997 Korea's Behavioral Risk Factor Surveilance System Survey collected through telephone interview based on the multi-stage strarified random sampling (N=1,055). For the analysis of the data on demographic and economic characteristics and health behavior, chi2 test and multiple logistic regression were used to determine the relationship between stress and smoking. RESULTS: Multiple logistic analysis showed that the group with mid level stress had a 1.52 times (95% CI=1.03~2.24) greater risk for smoking than the low stress group and the high stress group, a 2.34 times (95% CI=1.14~4.82) greater risk. There was a higher risk in men than in women, in the group who had experienced divorce or lost of spouse than for those who were married, in those who are employed than unemployed and in drinkers than abstainers. CONCLUSION: This study found that the group who experienced higher stress levels carried higher proportion of smokers. Therefore, sound stress management and development and implementation of systematic stress control programs will contribute to health promotion and disease prevention.