Intolerance of uncertainty (IU) is a multidimensional construct that has been proposed as an important transdiagnostic risk factor across various anxiety and mood disorders. Recent work found support for IU having a continuous latent structure when utilizing taxometric methods. However, taxometrics may not be ideally suited to examine the latent structure of constructs such as IU given the methodological shortcomings associated with this technique. The current study applied factor mixture modeling, a statistical technique that overcomes shortcomings of prior work, to examine the latent structure of IU in a sample of 371 individuals presenting at an outpatient clinic. Findings indicated that the best fitting solution was a 3-class model with 1 class consisting of individuals with high levels of IU (High IU; n = 55) and 1 containing individuals with low levels of IU (Low IU; n = 206). Our third class, labeled Moderate IU, consisted of 110 individuals with levels of IU between those of the High IU and Low IU groups. There were also significant differences across the 3 IU classes, including the relations between IU classes and anxiety-related and depressive disorders. The current investigation was the first to find evidence of IU having a categorical latent structure. Implications for research and clinical utility are discussed.