Host-bacterium interaction in experimental pseudotuberculosis infection in guinea pigs was investigated. Quantitation of the pyogranulomatous response induced at the infection site revealed a biphasic (perhaps triphasic) pattern. The first peak observed within 5-9 days after infection (depending upon the dose) was attributed to a non specific response of the body while the second peak seen 17-21 days postinfection might be due to a specific antibacterial immunity. The responses induced at the primary and challenge sites could be modulated by various treatments including altering the dose and timing of the challenge. Reinfection occurred , although at reduced level, when previously infected guinea pigs were challenged at about the time of the second peak of granulomatous response. Immunity was evidenced by apparent failure of the challenge organisms (a double antibiotic resistant mutant of the parent strain) to reach the secondary infection sites (local lymph nodes and internal organs) due either to failure of dissemination or prompt elimination because of the ensuing immunity. The present study gave evidence that administration of C. pseudotuberculosis in guinea pigs results in a systemic, possibly bacteraemic spread of the organisms which is probably irrespective of the route of administration. It was difficult to demonstrate bacteraemia by blood or organ cultures except when animals were given very large doses (10[9] cfu) by the i/p route. In rabbits following administration of a large dose by the i/p route two peaks were observed: the first occurred within 30 minutes while the second was observed between 30-60 minutes. No organisms could be detected after 2 hours. Bacteraemia almost certainly occurred in guinea pigs following oral infection, as evidenced by the localization of the organisms at sites inoculated with tissue-damaging substances (calcium chloride, powdered glass/sand mixture, incomplete Freund's adjuvant and Tween saline) within 2 hr of oral infection. The method described for inducing secondary localization of C pseudotuberculosis in the present study could be adopted as a model for other bacterial diseases where metastatic lesions are important in pathogenesis. Immune responses to C. pseudotuberculosis infection in guinea pigs as observed in the present study are discussed with reference to other facultative intracellular bacteria of veterinary importance. Implications of the findings in natural disease of sheep and goats are discussed with suggestions about the future directions such studies might take.