OBJECTIVES: As smoking is very common in R-rated films, we sought to determine if viewing R-rated films is associated with adolescent smoking.

METHODS: Three annual cross-sectional surveys conducted of 88,505 Year 10 students of largely European, Maori, Asian or Pacific Islander ethnicity in secondary schools in New Zealand between 2002 and 2004. Outcomes of interest were: intention to smoke among never smokers; past experimentation with smoking among current non-smokers; current smoking status; and current frequency of smoking.

RESULTS: Dose-response relationships were observed between the frequency of viewing R-rated films and all outcome measures controlling for age, gender, ethnicity, peer smoking, parental smoking, socioeconomic status, pocket money and household smoking rules. Compared to never viewing R-rated films, viewing at least weekly nearly tripled the relative risk (2.81; 95% confidence interval 2.57, 3.09) of never smokers being susceptible to smoking, and more than doubled the risk of both past experimentation (2.28; 95% CI 2.12, 2.45) and smoking > /=monthly (2.31; 95% CI 2.10, 2.54). Each of these risks was seen across all ethnic groups.

CONCLUSIONS: Our results extend the association that has been demonstrated between viewing R-rated films and current smoking in American youth by demonstrating the same association in youth of different ethnic and cultural backgrounds in New Zealand.