From an eligible population of 9,943 casino employees, 6,067 volunteered to participate in this study. Of this sample, 1,176 provided data at 3 observation points approximately 12 months apart. Using the South Oaks Gambling Screen (SOGS; H. R. Lesieur & S. B. Blume, 1987) and the CAGE (J. A. Ewing, 1984) questionnaire, the authors prospectively examined the prevalence and patterns of alcohol and gambling problems among those employees. Among the casino employees with gambling and drinking problems, a segment displayed the capacity to diminish those problems even when the difficulties had reached disordered levels. The authors also examined the comorbidity of gambling and drinking as well as the relationships among changes in SOGS scores and CAGE scores and changes in demographic and biological variables. The women were more likely to decrease their problem-drinking scores, but not their gambling scores, when compared with the men. In addition, 2 key variables (i.e., disabling depression and dissatisfaction with one's personal life) emerged as predictors of transitions to healthier levels of disordered gambling. The authors cautiously suggest, in light of the results taken together,that more fluctuation is associated with gambling and drinking problems than previously thought and that the conventional wisdom about disordered gambling as "always progressive" needs reconsideration.