Primary plaque forming cells (PFC) are present in spleens of mice 150 days or more following an infection with Brucella abortus. The development of primary plaques in mice long after antigenic challenge is an uncommon phenomenon, unlike the plaque formation (PF) induced by a non-living antigen. The mechanism of this persistent PF has been now investigated in light of a prolonged persistence of the corresponding antigen in tissues. Living E. coli, inoculated in massive dose into mice, survived in their organs for a brief time, while concomitantly PFC disappeared by day sixteen. Infection with B. abortus, in contrast, induced persistent presence of bacteria in the organs of inoculated mice and stimulated long lasting plaque formation. Only direct plaques were found during all stages of infection. Repeated inoculations of dead B. abortus also induced continuous production of primary plaques, whereas an interval in supply of the antigen resulted in disappearance of PFC. Rifampin (40 mg/kg) eliminated bacteria from the treated mice, which resulted in the disappearance of primary PFC. It seems likely that long lasting PF in B. abortus infected mice is connected with a constant antigenic stimulus operating in the carrier state.