Background and Aims In 2003, the Netherlands introduced tobacco control policies, including bans on tobacco sales to minors, advertising and sponsoring and tobacco sales in government institutions. We examined the extent to which these policies were associated with a change in educational inequalities in adolescent smoking. Design Repeated cross-sectional survey. Setting The Netherlands, 1992-2011. Participants A total of 43527 14-19-year-old adolescents. Measurements Data were obtained from the national Youth Smoking Monitor. We used logistic regression analyses to model the immediate change in daily smoking prevalence in 2003, the trends and the changes in trends. Models included interactions between educational level (high versus low, based on the educational track of the respondent) and, respectively, period (after versus before 2003), time and timexperiod. Findings Before 2003 the smoking trend declined slightly, and the decline was comparable for students of both high and low educational levels. Immediately after tobacco policies were introduced, daily smoking prevalence dropped for the total population [regression coefficient () = -0.340, 95% confidence interval (CI) = -0.445; -0.236]. This drop was larger for high educational level compared to low educational level ( interaction = -0.400, 95% CI = -0.623; -0.176). After 2003, trends in educational inequalities in smoking stabilized. Conclusions Following the introduction of new tobacco control policies in the Netherlands in 2003, smoking prevalence rates decreased among adolescents of both higher and lower educational levels. However, socio-economic inequalities in adolescent smoking increased.