Infected breeder pigs and contaminated feed represent potential sources of Salmonella introduction to fattening pig herds and may thereby cause human infections acquired via consumption of contaminated pork. Modelling approaches such as quantitative microbial risk assessment could improve the design of strategies for control and tracing of Salmonella in the feed chain. However, the construction of such models requires a thorough understanding of the dynamics of the feed chain, including production processes, microbial processes and transport logistics. The present article illustrates a conceptual model of Salmonella in the pig feed chain and explores the possibilities for quantitative modelling including identifying major gaps in data. Information was collected from peer-reviewed scientific journals, official documents and reports and by means of interviews with experts from authorities and the feed industry. Data on prevalence of Salmonella in different parts of the feed chain are difficult to compare as observed prevalence may be biased by variations in sampling procedures as well as limitations of the detection methods. There are almost no data on numbers of Salmonella in commodities of the feed chain, which often makes it difficult to evaluate risks, intervention strategies and sampling plans in a quantitative manner. Tracing the source of Salmonella contamination is hampered by the risk of cross-contamination as well as various mixing and partitioning events along the supply chain, which sometimes makes it impossible to trace the origin of a lot back to a batch or producer. Available information points to contaminated feed materials, animal vectors and persistent contamination of production environments as important sources of Salmonella in feed production. Technological procedures such as hydrothermal or acid treatment can be used to control Salmonella in feed production. However, a large fraction of pig feed is produced without decontamination procedures. Prevention of recontamination and control of moisture throughout the chain are thus critical factors for controlling Salmonella in feed production. To verify successful control it is necessary to have monitoring strategies able to detect low levels of Salmonella heterogeneously distributed in large volumes of feed and feed material in bulk. Experience from monitoring programs and research investigations indicates that sampling of dust and sweepings from control points along the production line is an efficient strategy to gain an indication of Salmonella contamination.