Purpose: Reputation is conceptualised as the believed effects that any social agent (ranging from a person to a company to a country) can have. Food reputation is beliefs about the effects of food on its consumers. On the basis of a multidimensional construct for food reputation derived from qualitative and correlational studies, this paper aims to test four hypotheses about food reputation dimensions' effects on consumers' food choices. Design/methodology/approach: A multi-attribute, multi-step choice experiment was carried out using a "phased narrowing" procedure. The procedure is based on eight product choices, using four reputation dimensions as manipulated attributes (duration, identity-territoriality, social and environmental responsibility, psycho-physiological well-being); this is replicated on one drink and one food product. Findings: A pilot study (n=50) checked the manipulation of the four reputation dimensions. ANOVA (n=118) showed the impact of the manipulated reputation features in the food choice process, especially in the final decision-making phase. Originality/value: The results confirm that food reputation features impact consumer choice, detailing the relative importance of different reputation features according to choice phase, product category, and individual characteristics.