The bioaccessibility of eicosapentaenoic acid (EPA) in the forms of monoacylglycerol (EPA‐MAG), triacylglycerol (EPA‐TAG), and phospholipid (EPA‐PL) during gastrointestinal passage was compared in this study using a dynamic gastrointestinal model (TIM system). The TIM system simulated the average upper gastrointestinal tract conditions of healthy human adults after intake of a meal (fed state conditions). In this study, the three EPA‐rich oils were separately homogenized with full fat milk to obtain oil‐in‐water emulsions. Plain yogurt was added into the mixture at an emulsion/yogurt ratio of 4:1 (w/w) as the food matrix of the test products. The results show that the test meals containing EPA‐PL left the stomach compartment most efficiently in comparison with the gastric emptying of EPA‐MAG and EPA‐TAG. The PLs also showed a significantly (P < 0.05) higher bioaccessibility of EPA (75–80%) in comparison with MAG (30%) and TAG (38%). The better gastric emptying of EPA‐PL was likely related to the more stable emulsion of EPA‐PL in the test meal. EPA‐PL was delivered within the meal matrix into the duodenum instead of floating on the top of the test meal matrix. EPA‐MAG had the highest amount of EPA that did not leave the stomach (68% of the test meal). The results from this work indicate that EPA‐PL is a more effective form of EPA for a higher lipid bioaccessibility than MAG and TAG under the test conditions.