Endophthalmitis is a devastating infection that can cause blindness. Over half of Bacillus endophthalmitis cases result in significant loss of useful vision. Bacillus produces many virulence factors that may contribute to retinal damage and robust inflammation. We analyzed Bacillus immune inhibitor A (InhA) metalloproteases in the context of this disease, hypothesizing that InhAs contribute to Bacillus intraocular virulence and inflammation. We analyzed phenotypes and infectivity of wild type (WT), InhA1-deficient (ΔinhA1), InhA2-deficient (ΔinhA2), or InhA1, A2, and A3-deficient (ΔinhA1-3) Bacillus thuringiensis. In vitro analysis of growth, proteolysis, and cytotoxicity were compared. WT and InhA mutants were similarly cytotoxic to retinal cells. Mutants ΔinhA1 and ΔinhA2 entered log phase growth earlier than WT. Proteolysis by the ΔinhA1-3 mutant was decreased, but this strain grew similar to WT in vitro. Experimental endophthalmitis was initiated by intravitreally infecting C57BL/6J mice with 200 CFU of B. thuringiensis WT or InhA mutants. Eyes were analyzed for intraocular Bacillus and myeloperoxidase concentrations, retinal function loss, and gross histological changes. Eyes infected with ΔinhA1 or ΔinhA2 strains contained greater numbers of bacteria than eyes infected with WT throughout the infection course. Eyes infected with single mutants had inflammation and retinal function loss similar to eyes infected with WT strain. Eyes infected with ΔinhA1-3 cleared the infection. RT-PCR results suggested that there may be compensatory expression of the other InhAs in the single InhA mutant. These results indicate that together, the InhA metalloproteases contribute to the severity of infection and inflammation in Bacillus endophthalmitis.