This study compares kinetic parameters of Salmonella and surrogate Enterococcus faecium in mash broiler feed during thermal inactivation. Two-gram samples of mash broiler feed were added into a filtered sample bag and inoculated with nalidixic acid (NaL, 200 ppm) resistant S. Typhimurium or Enterococcus faecium, followed by vacuum-packaging and heating in a circulated thermal water bath at 75°, 85°, and 95°C for 0 to 180 s. Counts of bacterial survival were analyzed on tryptic soy agar and bile esculin agar plus 200 ppm of NaL. Microbial data and thermal kinetic parameters (n = 8, Global-Fit and United States Department of Agriculture [USDA]-Integrated-Predictive-Modeling-Program software) were analyzed by JMP software. Heating mash broiler feed at 75°, 85°, and 95°C decreased (P < 0.05) Salmonella cell counts by >6 log10CFU/g after 180, 60, and 50 s, respectively. Heating E. faecium in feed at 75°, 85°, and 95°C for 180, 120, and 70 s achieved reductions of 3, 6, and >6.5 log10CFU/g, respectively. D-values of linear, Weibull models, and z-value of Salmonella at 75°, 85°, and 95°C were 1.8 to 11.2, 4.2 to 21.8, and 28.6 s, respectively, which were lower (P < 0.05) than those of E. faecium (3.7-18.1, 8.5-34.4, and 34.1 s). Linear with Tail, Linear with Tail and Shoulder, and Weibull with tail equations revealed that E. faecium were more resistant (P < 0.05) to heat than Salmonella as shown by longer "Shoulder-time" (26.5 vs. 16.2 s) and greater "Tail" effect (4.4-4.5 vs. 2.5-2.6 log10CFU/g). Results clearly suggested that E. faecium can be used as a surrogate for Salmonella to validate thermal inactivation during feed manufacture.