While the association between gambling disorders and comorbid mental disorders has been extensively studied, only a few studies have used longitudinal data or evaluated the association across different levels of gambling behavior and specific gambling-related symptoms. In this study, longitudinal data from waves 1 and 2 of the National Epidemiological Survey on Alcohol and Related Conditions (NESARC) were used to determine whether different levels of gambling behavior and gambling-related symptoms were associated with the onset of psychiatric disorders. Although NESARC used DSM-IV diagnoses, for this study, the recently published DSM-5 diagnostic criteria for gambling disorder were used to group the NESARC respondents (N = 34,653) into three levels of gambling (gambling disorder, sub-threshold gambling disorder, and recreational gambling) and one non-gambling comparison group. Three years after the initial intake interview, compared to the non-gamblers, those reporting any gambling behavior at baseline were at increased risk to have any mood, anxiety, or substance use disorders (recreational gambling: adjusted odds ratio [AOR] = 1.16, 95% confidence interval [CI] = 1.10-1.23; sub-threshold gambling disorder: AOR 1.77, 95% CI 1.63-1.92; gambling disorder: AOR 2.51, 95% CI 1.83-3.46). Similar graded relationships were found for a number of specific disorders. In addition, multiple specific gambling-related symptoms were associated with comorbid disorders, possibly suggesting the interaction of different mechanisms linking gambling disorder and the onset of comorbid psychopathology. In conclusion, a graded or dose-response relationship exists between different levels of gambling and the onset of comorbid psychopathology. Among gambling groups, those with a gambling disorder were at the highest risk for the new onset of comorbid conditions and those with recreational gambling were at the lowest risk, while the risk among participants with sub-threshold gambling disorder fell between these two groups.