By feeding Atlantic salmon diets with 64% of the fish oil (FO) replaced by vegetable oil, and with decreasing fishmeal (FM) inclusion levels from 213, 178 and 143 g kg−1 (accumulated level during the seawater phase) in a full-scale experiment producing 3.1 thousand tonnes fish, no significant negative effects on fish performance, health and product quality were observed. All dietary groups showed, however, moderate intestinal inflammation. Reduced growth and feed efficiency were seen with decreasing fishmeal inclusion levels. Two dietary groups demonstrated net marine protein production, while none of the groups showed net fish production (FIFO ≥1.65) due to the equal low FO inclusion. High plant oil level gave lower fillet level of persistent organic pollutants (POPs) compared with the levels surveyed on the Norwegian market. The study gave predictable incorporation rates of essential n-3 long-chain fatty acids in the fillet. Cooked salmon fillet from all dietary groups showed minor differences in sensory quality. Based on the present full-scale production results, dietary FM inclusion down to 160 g kg−1 (accumulated) during the seawater phase, concurrent to replacing ~70% of the FO with a suitable plant oil, is not regarded to represent any risk to fish performance, health or quality.