Variation in responses to pathogens is influenced by exposure history, environment and the host's genetic status. We recently demonstrated that human leukocyte antigen class II allelic differences are a major determinant of the severity of invasive group A streptococcal (GAS) sepsis in humans. While in-depth controlled molecular studies on populations of genetically well-characterized humans are not feasible, it is now possible to exploit genetically diverse panels of recombinant inbred BXD mice to define genetic and environmental risk factors. Our goal in this study was to standardize the model and identify genetic and nongenetic covariates influencing invasive infection outcomes. Despite having common ancestors, the various BXD strains (n strains=33, n individuals=445) showed marked differences in survival. Mice from all strains developed bacteremia but exhibited considerable differences in disease severity, bacterial dissemination and mortality rates. Bacteremia and survival showed the expected negative correlation. Among nongenetic factors, age -- but not sex or weight -- was a significant predictor of survival (P=0.0005). To minimize nongenetic variability, we limited further analyses to mice aged 40-120 days and calculated a corrected relative survival index that reflects the number of days an animal survived post-infection normalized to all significant covariates. Genetic background (strain) was the most significant factor determining susceptibility (P< or =0.0001), thus underscoring the strong effect of host genetic variation in determining susceptibility to severe GAS sepsis. This model offers powerful unbiased forward genetics to map specific quantitative trait loci and networks of pathways modulating the severity of GAS sepsis.