This paper proposes a novel type-2 fuzzy ontology, including a type-2 food ontology and a type-2 FML-based ontology, for diet assessment. In addition, the paper presents a type-2 fuzzy diet assessment agent. In the proposed approach, firstly, the nutrition facts of various kinds of food are collected from the Internet and the convenience store. Next, the domain experts construct the type-2 fuzzy ontology, then the involved volunteers are requested to input eaten items. Finally, the proposed agent displays the diet assessment of the input eaten items based on the constructed type-2 fuzzy ontology. The experimental results show that the proposed approach is feasible for diet assessment.