© Springer International Publishing Switzerland 2017 ; Purpose: Rising rates of obesity have been recently associated to the novel concept of food addiction (FA). The Yale Food Addiction Scale (YFAS) is the most widely used measure for examining FA (1) and analysis of its reliability and validity is expected to facilitate empirical research on the construct. Here, we tested the psychometric properties of a Portuguese version of the YFAS (P-YFAS), establishing its factor structure, reliability and construct validity. Methods: Data were obtained from 468 Portuguese individuals, 278 sampled from non-clinical populations, and 190 among obese candidates for weight-loss surgery. A battery of self-report measures of eating behavior was applied. Results: Confirmatory factor analysis verified a one-factor structure with acceptable fit, with item analysis suggesting the need to eliminate item 24 from the P-YFAS. Internal consistency (KR-20 = .82) and test–retest stability were adequate. Correlation analyses supported convergent and divergent validity of the P-YFAS, particularly in the clinical sample. Both FA symptom count and diagnosis, according to the P-YFAS, adequately discriminated between samples, with classification of FA met by 2.5 and 25.8% of the participants in the non-clinical and clinical samples, respectively. Conclusions: These findings reinforce the use of P-YFAS in non-clinical and clinical populations. Future directions for extending YFAS validation are discussed. ; This work was supported by Fundação para a Ciência e Tecnologia through a Junior Research and Career Development Award from the Harvard Medical Portugal Program (HMSP/ICJ/0020/2011) to AJO-M and CPUP (UID/PSI/00050/2013) to ST, and a Grant from the BIAL Foundation (176/10) to AJO-M. ; info:eu-repo/semantics/publishedVersion