Mesoporous silica (MPS) nanoparticles (NPs), which have a unique pore structure and extremely large surface area and pore volume, have received much attention because of their biomedical application potential. Using MPS NPs for biomedical devices requires the verification of their biocompatibility because the surface area of NPs is one of the most important determinants of toxicity, including the cellular uptake and immune response. We have previously reported that the cytotoxicity and inflammation potential of MPS NPs have been shown to be lower than those of general amorphous colloidal silica (Col) NPs in macrophages, but the low cytotoxicity does not guarantee high biocompatibility in vivo. In this study, we compared the in vivo immunotoxicity of MPS and Col NPs in the mouse model to define the effects of pore structural conditions of silica NPs.Both MPS and Col NPs (2, 20, and 50 mg/kg/day) were intraperitoneally administered in female BALB/c mice for 4 weeks, and clinical toxicity, lymphocyte population, serum IgG/IgM levels, and histological changes were examined.There was no overt sign of clinical toxicity in either MPS- or Col-treated mice. However, MPS NPs led to significant increases in liver and spleen weight and splenocyte proliferation. Mice treated with MPS NPs showed altered lymphocyte populations (CD3(+), CD45(+), CD4(+), and CD8(+)) in the spleen, increased serum IgG and IgM levels, and histological changes. Despite slight changes in lymphocyte populations in the spleen, Col NPs did not alter other immunological factors.The results indicate that in vivo exposure to MPS NPs caused more damage to systemic immunity than that of Col NPs through the dysregulation of the spleen. The results for in vivo data are inconsistent with those for in vitro data, which show lower cytotoxicity for MPS NPs. These results suggest the importance of verifying biocompatibility both in vitro and in vivo during the design of new nanomaterials.