The aim of the study was to test the effectiveness of a postpartum smoking cessation and relapse prevention intervention. Structural equation modeling techniques were applied to evaluate the impact of the intervention on smoking behavior and on non-behavioral variables derived from the Transtheoretical Model (TTM). Women were randomized to an intervention (I) and control group (C). Smoking status, TTM-variables, and control variables were assessed four weeks, six and 12 months postpartum. Membership in the intervention group significantly predicted non-smoking and higher self-efficacy six months, but not one year postpartum, after controlling for demographic, smoking, and postpartum risk variables.