Introduction: Cannabis is one of the most widely used psychoactive substances in the United States (U.S.). Perceived risk of use is associated with substance use; the recent debate surrounding medicalization and legalization of cannabis in the U.S. has the potential to impact perceived risk of use. Recent estimates are needed to assess temporal changes in, and identify correlates of, perceived risk of cannabis use. Methods: Utilizing data from the 2002–2012 survey years of the National Survey on Drug Use and Health, chi-squared statistics and logistic regression were used to describe temporal changes in perceived risk of regular cannabis use (i.e., once or twice a week), to explore correlates of perceived risk, and to report frequency of cannabis use. Results: Between 2002 and 2012, perceived great risk of regular cannabis use varied significantly overall (p < 0.001). The prevalence of past year non-daily (p < 0.001) and daily use varied significantly during this time (p < 0.001). Controlling for survey year and other confounders, characteristics associated with increased odds of perceived great risk of regular cannabis use included: female sex; Non-White race/ethnicity; age 50+; and family income of $20,000-49,999. Characteristics associated with decreased odds of perceived great risk included: ages 12–17 and 18–25; high school education or greater; total family income of $75,000+; past year non-daily and daily cannabis use; and survey years 2008–2012. Conclusions: Findings characterize trends of perceived risk of regular cannabis use, and past year non-daily and daily cannabis use. Longitudinal studies of the influence of legal status of cannabis at the state-level are needed.