Metallic biomaterials are generally used for replacement of structural components of the human body such as bones, joints, and tooth roots. When they are implanted inside a body, metallic biomaterials may corrode and/or wear, releasing metal ions and debris which may have toxic effects on tissues and organs. Since it is important for biomaterials to have no toxicity against a living body, a systematic and quantitative evaluation of the cytotoxicity of metallic elements is required for the development of new metallic biomaterials with superior biocompatibility. In this study, the cytotoxicity of 43 metal salts were evaluated by the colony formation method using two kinds of cultured cells. The effects of the difference in valence numbers of metallic elements in the salts on cytotoxicity were examined. The cytotoxicity of the salts of metallic elements' oxo acids was also investigated. As a result, the intensity of metal salts' cytotoxicity tends to be quite similar between MC3T3-E1 and L929 (the correlation coefficient of metal salts' IC50s is 0.82). The intensity of metal salts' cytotoxicity depends on the kinds of metallic elements, their chemical states, and concentrations. The IC50 of the highest toxic salt is 1.36 x 10(-6) mol L-1, which differs four orders of magnitude from the IC50 of the lowest toxic salt. K2Cr2O7, CdCl2, VCl3, AgNO3, HgCl2, SbCl3, BeSO4, and InCl3 are high toxic salts in which IC50s are smaller then 10(-5) mol L-1 for both or either of the cell lines. HgCl, Tl(NO3)3, GaCl3, CuCl2, MnCl2, CoCl2, ZnCl2, NiCl2, SnCl2, IrCl4, TlNO3, CuCl, RhCl3, Pb(NO3)2, Cr(NO3)3 and Bi(NO3)3 are relatively high toxic salts in which IC50s are smaller than 10(-4) mol L-1 for both or either cell lines.