This study examines, among a large health maintenance organization population, the prevalence of two high-risk lifestyle practices (smoking and problem drinking), their interrelationships, and their relationships with other lifestyle practices, sociodemographic characteristics, and health status measures. Results, based on a random sample of 1,133 adults, showed that smoking and problem drinking are strongly correlated. Individuals with no drinking problems had an age-, sex-, and education-adjusted smoking prevalence of approximately 20%, while problem drinkers smoked at about twice that rate. In addition, reporting one type of problem drinking behavior (binge, chronic, or drinking and driving) at least doubled, and in one instance increased by sixfold, the likelihood of reporting another type of problem drinking behavior. Smokers and problem drinkers were more likely to be younger than age 65, to be irregular seat belt users (smokers and binge drinkers only), and not to belong to voluntary organizations. Results of the analysis suggest that detection, prevention, and treatment of drug use, in general, might prove more beneficial than only focusing on smoking and problem drinking. In addition, because binge drinking and drinking and driving were so widespread among younger age groups, it might prove more beneficial to consider preventive strategies that change the sale and distribution of alcohol and make the environment safer in which to drink, such as providing transportation to get drinkers back home.