Objective. To extend the quitting continuum to categorize smokers in the early stages of the process of smoking cessation. Design. A prospective computer assisted telephone interview study with initial interviews in 1990 and reinterviews in 1992. Setting and participants. In California, 2514 current smokers and quitters of less than 5 years duration, selected randomly from a large scale cross-sectional survey in 1990. Results. We focus on smokers with high addiction (15 cigarettes/day) and without a strong quitting history (1 week in last year or 1 year ever) at baseline (N=822). Having an intention to quit and a limited quitting history (1-6 days in last year) was predictive of progress at follow-up into higher continuum levels ultimately associated with successful cessation. Furthermore, smokers with one trait were much more likely to progress than those with none. Accordingly, the lowest level of the earlier continuum was further subdivided into three subgroups (making eight levels overall). Over approximately a 2-year period, most smokers either progressed or regressed only one or two levels along the quitting continuum. Conclusions. Better than expected progress along this expanded quitting continuum could be a criterion for a successful intervention. Tailored interventions that move smokers to a higher level on the continuum should be a priority for future research.