This paper addresses the problem of assessing nicotine dependence. The main objective is to develop theory-led suggestions for measures that will be relevant in the early phases of tobacco use, as well as in established smokers. Theoretical models of addiction falling into the general class of 'positive reinforcement theories' were identified and reviewed. From this review a number of drug effects and patterns of behaviour were distilled and categorized as either vulnerability or dependence indicators. A comparison of those features with the International Classification of Diseases (ICD-10) and Diagnostic and Statistical Manual (DSM-IV) diagnostic systems shows that neither system includes detailed assessment of vulnerability indicators. It is argued that measurement of vulnerability indicators, in addition to dependence indicators, may add to the predictive validity of assessments carried out in early career tobacco users, especially where there is limited evidence of established dependence. In addition, it is suggested that examination of measures that differentiate a subgroup of early career smokers termed 'rapid accelerators' may prove profitable and enable identification of the key parameters of nicotine reinforcement.