ABSTRACT: BACKGROUND:Tularemia is caused by Francisella tularensis, a gram-negative bacterium that has been weaponized as an aerosol. For protection of personnel conducting biodefense research, the United States Army required clinical evaluation of a new lot of tularemia live vaccine strain manufactured in accordance with Current Good Manufacturing Practices. METHODS:A phase 2 randomized clinical trial compared the new lot (DVC-LVS) to the existing vaccine that has been in use for decades (USAMRIID-LVS). The vaccines were delivered by scarification to 228 participants. Safety, reactogenicity, take and/or antibody levels were assessed on days 0, 1, 2, 8, 14, 28, 56, and 180. PRINCIPAL RESULTS:Both vaccines were safe and had acceptable reactogenicity profiles during six months of follow-up. There were no serious or grade 3 and 4 laboratory adverse events. Moderate systemic reactogenicity (mostly headache or feeling tired) was reported by ?23% of participants receiving either vaccine. Injection site reactogenicity was mostly mild itchiness and pain. The frequencies of vaccine take skin reactions were 73% (95% CI, 64, 81) for DVC-LVS and 80% (95% CI, 71, 87) for USAMRIID-LVS. The 90% CI for the difference in proportions was -6.9% (-16.4, 2.6). The rates of seroconversion measured by microagglutination assay on days 28 or 56 were 94% (95% CI, 88, 98; n=98/104) for DVC-LVS and 94% (95% CI, 87, 97; n=103/110) for USAMRIID-LVS (p=1.00). Day 14 sera revealed more rapid seroconversion for DVC-LVS relative to USAMRIID-LVS: 82% (95% CI, 73, 89) versus 55% (95% CI, 45, 65), respectively (p<0.0001). MAJOR CONCLUSIONS:The DVC-LVS vaccine had similar safety, reactogenicity, take and antibody responses compared to the older USAMRIID vaccine, and was superior for early (day 14) antibody production. Vaccination take was not a sensitive surrogate for seroconversion in a multi-center study where personnel at five research clinics performed assessments. ClinicalTrials.gov identifier NCT01150695.