Objectives. The present study examined the ability of the TPB to predict the intention to quit smoking and quitting behaviour. In addition, the predictive power of future orientation, number of cigarettes smoked, planning, past behaviour and the interactions between intention and other predictors was examined.

Material and methods. The data were derived from a longitudinal survey among 103 daily smoking students at the University of Oslo (mean age 24.6 years, mean number of years of smoking = 8 years). These data were collected by means of self-administered questionnaires at T1 (October 2003) and at T2 (February 2004) in terms of recording actual quitting.

Results. The TPB components accounted for 30% of the variance in quitting intentions, and affective attitude and descriptive norm emerged as the strongest predictors of quitting intention. Ordinal regression analysis showed that intention was a borderline significant predictor of subsequent quitting behaviour, while the impact of PBC was non-significant (model 1). The inclusion of the additional variables improved the fit of the model, with number of cigarettes and planning appearing as significant predictors of behaviour (model 2). As predicted, there was a significant interaction between perceived control and intentions on quitting (model 3). Nagelkerke R2 increased from .07 in model 1 to .54 in model 2, and finally to .58 in model 3.

Conclusions. The results indicate that affective attitude and descriptive norm play a more crucial role than the other TPB predictors in motivating smokers to quit. The results also indicate that self-regulatory strategies are important in relation to addictive behaviours.