Objective: The relationship between the level of gambling fallacy endorsement and type of gambler (nongambler, recreational gambler, at-risk gambler, and problem/pathological gambler) was assessed both concurrently and prospectively in a large national cohort of Canadian adults. Method: This cohort (n = 10,199 at baseline; 18-24 years, n = 481, 43% female; 25-34 years, n = 1,335, 62% female; 35-44 years, n = 1,543, 55% female, 45-54 years, n = 1,985, 58% female; 55-64 years, n = 2,459, 55% female; 65-74 years, n = 1,865, 44% female, 75+ years, n = 531, 43% female) was recruited from LEO, Leger Opinion's registered online panelists. The follow-up survey was completed by 55.9% of the cohort, 1 year after baseline. The full survey can be viewed at https://www.ucalgary.ca/research/national-gambling-study/. For the current study, scores on the Gambling Fallacies Measure, the Problem and Pathological Gambling Measure, Gambling Participation Instrument, and Impulsivity were analyzed. Results: There were three main findings. The first is that gambling fallacies are common in all categories of gamblers but somewhat more prevalent in problem and pathological gamblers. Second, the multivariate analysis determined that gambling fallacies are significant concurrent and prospective predictors of the problem/pathological gambling category, but not strong predictors relative to other variables. Third, problem gambling and heavier gambling involvement are also predictors of a future higher level of gambling fallacies. Conclusions: Collectively, these results show that gambling fallacies have some etiological relationship to problem gambling but are not the main cause of problem gambling and should not be the exclusive focus of problem gambling treatment. (PsycInfo Database Record (c) 2021 APA, all rights reserved).