Current smoking cessation treatments seem to differ from one another in the proportion of ex-smokers who survive the first few days of the quit attempt. After this initial effect, parallel relapse processes appear to unfold in all treatment groups; no available treatments seem to alter the nature of this late relapse process. True relapse prevention will require that we obtain a better understanding of the forces contributing to relapse across the span of the cessation attempt. A working model of dynamic relapse processes may be necessary before treatments can be improved. In this paper, we suggest that the conceptual model of relapse proneness (RP) described originally by the National Working Conference on Smoking Relapse can serve as an ecumenical organizational framework that may be used to integrate and conceptualize relapse data in ways that could generate new strategies for relapse research and inform treatment design. As an illustration, we sketch a preliminary model of RP which postulates that physical withdrawal, stressors/temptations, and cessation fatigue each make independent, time-shifted contributions to relapse risk. A new round of descriptive research focused on relapse proneness processes may be a prerequisite for improving existing smoking cessation interventions.