Food addiction (FA) has been examined in different populations. Although high FA levels are associated with greater eating disorder severity, few studies have addressed how FA relates to treatment outcome. Goals The study aims (1) to determine whether a brief intervention for bulimia nervosa (BN) reduces FA diagnosis or severity compared with baseline and (2) to determine if FA is predictive of treatment outcome. Method Sixty‐six female BN patients participated in the study. The Yale Food Addiction Scale was administered at two time points: prior to and following a 6‐week intervention. The number of weekly binging/purging episodes, dropout and abstinence from bulimic behaviour were used as primary outcome measures. Results This brief intervention reduced FA severity and FA diagnosis in the 55 patients who completed treatment. FA severity was a short‐term predictor of abstinence from binging/purging episodes after treatment (p = .018). Conclusions Food addiction appears to be prevalent in BN although FA severity can improve following a short‐term intervention.