Two demographically similar rural counties in central Virginia were compared, one intervention and one control site. In each country, two cohorts were assembled, a random sample of all Black smokers (the community cohort), and all members of randomly sampled Black churches (the church cohort). The descriptive component of the study elicited the knowledge, beliefs, attitudes, and behavior concerning smoking from these cohorts. Cardiovascular risk factors, including blood pressure and cholesterol levels, were also determined in the community cohort. For the experimental component, a variety of interventions which have proven to be efficacious in other settings were implemented through Black churches. These included fostering a supportive environment, distributing self-help materials, and providing individual support and counselling through the use of peer facilitators. The results of the baseline psychosocial survey were used to modify these methods for local use. They were then implemented in a culturally appropriate fashion at the intervention sites. Baseline smoking rates were determined using the initial surveys in both counties. Smokers were surveyed again at the end of the interventions and one year later to establish one-year abstinence rates. These were used as the primary outcome variables for the analysis of effect. The principal effect of the interventions were measured by comparing smoking abstinence rates between the church cohorts in the intervention and control counties. Due to the extended nature of the social networks associated with the churches, the interventions may also have had an effect on individuals not directly exposed, and this was determined by comparing the outcomes in the community cohorts. Finally, the associations between the baseline psychosocial and physiologic variables, exposure to the interventions, and smoking cessation behavior were also analyzed, to characterize on a population basis those who were and were not influenced by these church-based programs.