This study examined the degree to which different policy-based whole-school smoking actions were associated with cigarette smoking among 4,697 ninth-grade students in 31 schools in Perth, Australia. Smoking is banned in all schools in Western Australia. Cross-sectional self-report surveys measured student smoking prevalence and intensity and its correlates. Simultaneously, school policies and practices related to cigarette smoking were assessed by interviews with principals from each of the 31 schools. The relationship between school policies and practices and cigarette smoking were evaluated using logistic regression. After controlling for socioeconomic status, family smoking, and gender, no differences in “ever smoking” and “regular smoking” rates were seen among schools that reported having a school health policy, a school drug policy, or school health committee. Both ever smoking ( p <.001) and regular smoking ( p <.001) rates among students were lower at schools that provided education or counseling as well as discipline measures for students caught smoking at school, compared with schools that used discipline-only approaches. Although school health policies, drug policies, and health committees appeared to be associated with some psychosocial correlates of smoking, the associations of an education/counseling approach with these correlates was more consistent. These findings suggest that the actions taken to deal with students who violate smoking policy restrictions may be more important in reducing cigarette smoking than the presence of health or drug policies or health committees. Using education/counseling and discipline strategies rather than discipline only may help to reduce teenage smoking.