The objective of our experiments was to study the persistence and dissemination of orally administered Salmonella in smoltified Atlantic salmon. In experiment 1, salmon kept at 15°C were fed for 1 week with feed contaminated with 96 most-probable-number units of Salmonella Agona per 100 g of feed and then starved for 2 weeks. Samples were taken from the gastrointestinal tract and examined for Salmonella 1, 2, 8, 9, 15, and 16 days after the feeding ended. In experiment 2, Salmonella Agona and Montevideo were separately mixed with feed and administered by gastric intubation. Each fish received 1.0 × 108, 1.0 × 106, or 1.0 × 104 CFU. The different groups were kept in parallel at 5 and 15°C and observed for 4 weeks. Every week, three fish in each group were sacrificed, and samples were taken from the skin, the pooled internal organs, the muscle, and the gastrointestinal tract and examined for the presence of Salmonella. The results from the two experiments showed that the persistence of Salmonella in the fish was highly dependent on the dose administered. Salmonella was not recovered from any of the fish that were fed for 1 week with the lowest concentration of Salmonella. In the fish given the highest dose of Salmonella, bacteria persisted for at least 4 weeks in the gastrointestinal tract as well as, to some extent, the internal organs. The present study shows that under practical conditions in Norway, the risk of Salmonella in fish feed being passed on to the consumer of the fish is negligible.