Health-related behaviors are significant contributors to morbidity and mortality in the United States, yet the empirical evidence on the underlying causes of the vast within-population variation in health-related behaviors is mixed. While many potential causes of behaviors have been identified—such as schooling, genetics, and environments—little is known on how much of the variation across multiple health-related behaviors is due to a common set of causes. We use three separate datasets on U.S. twins to investigate the degree to which multiple health-related behaviors correlate and can be explained by a common set of factors. Based on the results of both within identical twin regressions and multivariate behavioral genetic models, we find that aside from smoking and drinking, most behaviors are not strongly correlated among individuals. While we find some evidence that schooling may be related to smoking, schooling is not a strong candidate explanation for the covariation between multiple behaviors. Similarly, we find that a large fraction of the variance in each of the behaviors is consistent with genetic factors; however, we do not find strong evidence that a single common set of genes explains variation in multiple behaviors. We find, however, that a large portion of the correlation between smoking and heavy drinking is consistent with common, mostly childhood, environments–suggesting that the initiation and patterns of these two behaviors might arise from a common childhood origin. Research and policy to identify and modify this source may provide a strong way to reduce the population health burden of smoking and heavy drinking.