In this work, it is shown that the common toxicity indicator, MTT (3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide), will fail to predict the toxicity of porous silicon (PSi) microparticles. This is due to the spontaneous redox reactions where the MTT is reduced and the PSi particle surfaces are oxidized simultaneously. MTT was shown to even react with thermally oxidized and carbonized forms of PSi particles, although the treatment did give an enhanced protection against the unwanted reactions as compared to as-anodized PSi particles. The observed levels of cellular viability with the MTT assay were much higher than expected in the presence of Caco-2 cells, even considering the spontaneous reduction of MTT at PSi surfaces. The results indicate that the redox reaction is further enhanced inside living cells. Thus, we recommend that MTT should not be used to test the cytotoxicity of drug formulations containing PSi microparticles. The study also shows that since PSi particles are capable of reducing the MTT, they will also be able to reduce other species as well. This should be taken into account when considering future applications for the porous silicon particles. The completely oxidized SiO2 particles (MCM-41 and SBA-15) were shown to work as expected with the MTT assay and showed no inherent oxidation/reduction.