Background Studies on the effects of school smoking policies are inconclusive and there is no research on whether the effects of school policies vary by educational level. We examined the association between school smoking policies and smoking behaviour among adolescents aged 14–17 years in Europe and assessed educational inequalities in these associations.

Methods Data on 10 325 adolescents from 50 schools in six European cities were obtained from the 2013 SILNE survey. We measured student perceived policy, staff reported policy and its three subscales: regulations, communication and sanctions. The association between school policies and smoking outcomes (daily smoking and smoking on school premises) was adjusted for individual characteristics and for parental smoking. We tested interaction between school policies and educational level.

Results Daily smoking was not associated with school smoking policies (eg, OR total policy=1.04, 95% CI 0.93 to 1.16 and OR student perceived policy=1.04, 95% CI 0.98 to 1.10). Smoking on school premises was less prevalent in schools with stronger staff reported total policy (OR=0.71, 95% CI 0.53 to 0.96). Other policy variables were also negatively associated with smoking on school premises, but not significantly (eg, OR student perceived policy=0.89, 95% CI 0.78 to 1.02). Associations between policy and smoking on school premises tended to be stronger in those with a low educational level, but none of the interactions tested were statistically significant.

Conclusions Our results suggest that school smoking policies may not have a direct effect on daily smoking but may reduce smoking on the school premises. We found no clear evidence for the effects of school policies to differ by educational level.