The Yale Food Addiction Scale (YFAS) is a widely used tool to assess the behavioural indicators of addictive-like eating. No studies, however, have used a longitudinal design to determine whether these addictive-like eating behaviours are a stable or transient phenomenon in a community-based population. This study aimed to evaluate whether food addiction Diagnosis and Symptom scores as assessed by the YFAS remain stable over 18-months in a non-clinical population. Young adults aged 18–35 years were recruited from the community to a web-based survey in 2013. The survey included demographics, anthropometrics and the YFAS. Participants who volunteered to be recontacted for future research were invited to complete the same survey 18-months later. The YFAS scoring outputs Diagnosis and Symptom scores were tested for agreement and reliability between the two time points. Of the 303 participants who completed the original survey and agreed to be recontacted, 69 participants (22.8% of those recontacted, 94% female, 67% normal weight at baseline) completed the 18-month follow-up survey. At baseline, thirteen participants met the YFAS predefined criteria for Diagnosis, while eleven participants met these criteria at the 18-month follow-up. YFAS Diagnosis was found to have moderate agreement [K = .50, 95% CI (.23, .77)] between the two time points while Symptom scores had good agreement [K = .70, 95% CI (.54, .83)]. Intraclass correlation coefficients were interpreted as moderate over the 18-month period for both the Diagnosis [ICC = .71, 95% CI (.45, .88)] and Symptom scores [ICC = .72, 95% CI (.58, .82)]. YFAS assessed food addiction Diagnosis and Symptom scores were found to be relatively stable over 18-months in a non-clinical population of predominantly female, young adults. Future research is required to determine the impact of behavioural weight loss interventions on YFAS assessed addictive-like eating.