The variable fecal shedding of Salmonella enterica serovar Typhimurium is an important cause of foodborne illness. Detection of Salmonella-carrier pigs is often difficult, as colonized pigs are often asymptomatic. To understand the transcriptomic response associated with the carrier state, as well as to determine if classifiers for carrier animals can be successfully built, we investigated the peripheral blood transcriptome following Salmonella inoculation in pigs differing in their Salmonella shedding phenotypes in two challenge populations (n=40 and 77). Salmonella inoculation introduced substantial gene expression changes at 2 days post inoculation (dpi) compared to 0 dpi. Differential expression of many genes between the two population extremes for shedding, termed low shedding (LS) and persistently shedding (PS) pigs were also observed. Analysis of the d2 dpi/d0 dpi profiling data identified distinct regulatory cascades mediated by IFN-γ, NF-kB, and several miRNAs, providing insight into the transcriptomic signals associated with distinct Salmonella fecal shedding phenotypes that can be targeted to further define why some animals are more resistant to Salmonella. To build transcriptome-based classifiers for distinguishing LS from PS pigs, a computational approach was applied. A class discovery procedure discovered discriminative patterns between LS and PS pigs by using the shedding data as prior knowledge. A class predictor of approximately 100 genes was then able to predict the shedding class of a second challenge population. Together, these results demonstrate the feasibility of Salmonella resistance classification based on transcriptomic data and suggest a general strategy for discovering and predicting classes for other bacterial pathogens.