Water pollution in aquaculture occurs mainly due to fish feces and uneaten feed. Supplying excess food to fish does not accelerate their growth; instead, this excess can cause issues, such as reduced digestion efficiency and increased water pollution. Therefore, an appropriate feeding frequency must be selected to optimize fish growth and minimize negative impacts. Matching the feeding frequency with peak appetite may enhance growth, improve feed conversion efficiency, and minimize water pollution. The return of fish appetite is associated with the gastric evacuation rate. The gastric evacuation of fish is affected by several factors, including fish size. To date, only a few studies have been conducted on the effects of the size of the Atlantic salmon (Salmo salar), an ecologically and culturally important salmonid fish, on the gastric evacuation rates in this species. Therefore, we aimed to determine the gastric evacuation rates and associated mathematical models of Atlantic salmon of different sizes and provide theoretical information for the feeding frequency in the context of aquaculture practice. The gastric evacuation experiment using two sizes of Atlantic salmon (176.15±27.52 g and 323.33±43.91 g) was carried out by analyzing the gastric contents. After adapting to the experimental environment for 27 days and fasting for 48 h, the wet and dry masses of the stomach contents of Atlantic salmon were assessed at predetermined postprandial times (0, 3, 6, 9, 12, 18, and 24 h) via serial slaughtering. Thereafter, the relative wet and dry masses at different time points were fitted using four mathematical models (linear, exponential, square root, and logistic models). The gastric evacuation time for 80% dry mass was calculated as the optimal feeding frequency after selecting the optimal model for each size. The water content in Atlantic salmon increased significantly from 0 to 3 h after feeding and then tended to be stable. Due to the moisture in the stomach, gastric evacuation of the wet mass displayed a different trend from that of the dry mass in both sizes of salmon. The dry-mass fitting model could reflect the gastric evacuation more accurately than the wet-mass fitting model. The optimal dry mass gastric evacuation model of small- and large-size Atlantic salmon aligned with the exponential model (reflecting a trend of fast and then slow) and square root model (reflecting a trend of slow and then fast), respectively. The gastric evacuation times of 80% dry mass in small- and large-size Atlantic salmon were 12.23 and 18.06 h, respectively, which indicated that the small-size Atlantic salmon had faster gastric evacuation time than the large-size salmon. In this study, a “lag phase” was observed within 3 h postprandial. Further, the “lag phase” of the small-sized salmon was found to be shorter than that of the large-sized salmon, which caused differences between the gastric evacuation models for the two salmon sizes. The small-sized salmon may have a faster gastric evacuation rate as softening of the dry feed occurs faster with water during the early digestion of the smallsized Atlantic salmon. These results could provide biological parameters and theoretical references for an innovative feeding strategy applicable to the production of Atlantic salmon.