BACKGROUND Most clinical trials evaluating treatments for alcohol use target individuals meeting diagnostic criteria for alcohol use disorder (AUD), but few address change in diagnostic status following treatment or as a potential outcome indicator. This study evaluated whether DSM-5 AUD total criteria count or severity category was sensitive to change over time and treatment effects. METHODS Data were drawn from a randomized clinical trial that evaluated the efficacy of computer-based cognitive behavioral therapy program (CBT4CBT) for AUD. Sixty-eight individuals were randomized to 1 of the 3 weekly outpatient treatments for an 8-week period: (i) treatment as usual (TAU), (ii) TAU+CBT4CBT, and (iii) CBT4CBT+brief monitoring. Structured clinical interviews were used to determine current (past 30 days) AUD diagnosis at baseline, end-of-treatment, and 6 months following end-of-treatment. Change in the total number of DSM criteria endorsed, as well as severity categories (mild, moderate, severe), was evaluated across time and by treatment condition. RESULTS Generalized Poisson's linear mixed models revealed a significant reduction in the number of DSM criteria from baseline to treatment end point [time effect χ2 (1) = 35.54, p