In experimental immunology, it frequently is of interest to obtain quantitative estimates of the virulence of strains of microorganisms as evidenced by their ability to produce fatal infections in experimental animals. It is recognized that, in the broader sense, virulence is not a specific intrinsic attribute of an organism but is the result of a combination of a number of factors pertaining to the organism, the affected host, and the conditions (known or unknown) under which the pathological state is established. Nonetheless, when suitable laboratory animals are employed, and when the combination of experimental conditions is propitious, it can be demonstrated with some consistency that even closely related strains of organisms differ in their capability of producing experimental infections. For the purpose of this presentation, such differences in capability can be considered as differences in virulence of the organisms regardless of the true nature of the experimental infective process.Quantitative estimates of virulence have been employed by various workers in studying the relationship of this property to antigenic constitution, bacterial dissociation, and various epidemiological phenomena. Furthermore, it popularly is considered that the immunogenic activity of a strain of bacteria is directly related to its virulence. As a consequence, such tests are employed extensively as a control technique in routine production of bacterial vaccines and in searching for immunogenically superior strains.Mice have been widely accepted as the laboratory animals of choice for determining the virulence of strains of Salmonella typhosa and the testing procedures employed have been of two main types. The first, based on the intraperitoneal injection of graded doses of the organisms suspended in physiological saline, is utilized almost exclusively by British workers. The second, incorporating the use of hog gastric mucin as a suspending medium, has been employed by most American workers since the original reports of the technique by Nungester et al.(1) and of its application to typhoid by Rake (2). A third method, which was reported by Norton and Dingle (3) to show promise, is based on the intracerebral injection of the organisms without the aid of an adjuvant. This latter method has been used successfully in determining the virulence of strains of Hemophilus pertussis by Kendrick and her coworkers (4, 5) and as an experimental infective technique with Shigella dysenteriae by Dubos et al.