Non-pathogenic lactic acid bacteria (LAB) are attractive as live carriers to deliver protective antigens to the mucosal immune system. Both persisting and non-persisting strains of lactic acid bacteria have been evaluated and seem to work equally well by the systemic and nasal routes of administration. However, it is not known if persistence and viability of the strain play a critical role when immunizing by the oral route. To address this question, recombinant LAB strains, able to persist (Lactobacillus plantarum NCIMB8826/pMEC127) or not (Lactococcus lactis MG1363/pMEC46) in the gastro-intestinal tract of mice and producing equivalent amounts of the tetanus toxin fragment C (TTFC) were compared to each other. A very strong ELISA TTFC-specific and protective humoral response was elicited by either live or UV-inactivated recombinant Lb. plantarum strains. In a similar protocol, recombinant Lc. lactis seemed to be somewhat less efficient than the former host. It is thus tempting to propose that the difference in the capacity of the bacterial vector to persist in the gastro-intestinal tract impacts on its immunogenicity and on the level of protection it may induce. Protection was slightly superior after administration of live strains.