Cigarette smoking is a major cause of cancer disability and deaths in the U.S. There is a significant need for developing methodologies to obtain representative, cost-effective samples of persons (especially siblings) for genetic studies on cigarette smoking. We propose to study the feasibility of using Random-digit dialing (RDD) computer-assisted telephone interview (CATI) surveys to create an adult sample of current smokers, former smokers, and never smokers who will also provide buccal cell DNA samples by mail. The innovativeness derives from using RDD telephone surveys to link self-reports of smoking behavior to DNA. Differential cooperation rates will be estimated for a study will be "made anonymous" with no follow-up versus a study that will be trying to establish a registry for confidential follow-up research with the same individual. Biases in non-response will be estimated by exploring what factors (e.g., sex, age, smoking status, education, geographic region) predict non-response at subsequent levels. Finally, genetic analyses will be performed to assess effects of candidate genes (CYP2A6 and CYP2D6) related to smoking phenotypes (smoking status, heaviness of smoking, and sensitivity to the first cigarette of the day). This is a multi-disciplinary proposal involving the techniques of social science as well as of molecular genetics.