Described an examination of data collected 2 years following the onset of a media-based, worksite smoking cessation intervention. Thirty-eight companies in Chicago were randomly assigned to one of two experimental conditions. In the initial 3-week phase, all participants in both conditions received self-help manuals and were instructed to watch a 20-day televised series designed to accompany the manual. In addition, participants in the group (G) condition received six sessions emphasizing quitting techniques and social support. In the second phase, which continued for 12 months, employees in G participated in monthly peer-led support groups and received incentives, while participants in the nongroup (NG) condition received no further treatment. Twenty-four months after pretest, 30% of employees in G were abstinent compared to only 19.5% in NG. This study is one of the few experimentally controlled worksite smoking cessation interventions to demonstrate significant program differences 2 years following the initial intervention.