Pretreatment of mice with cobaltous chloride protects them against a potentially lethal dose of sodium cyanide or sodium sulfide. The mechanism of this protective effect is unknown but is presumed to reflect a direct chemical reaction between cobalt and cyanide or sulfide. Methemoglobinemia was not observed after protective doses of cobaltous chloride in mice. Pretreatment of mice with sodium cobaltinitrite also protects them against potentially lethal doses of sodium cyanide or sulfide. At least part of this protective mechanism, however, must be ascribed to methemoglobin formation by the injected cobaltinitrite. A comparison of cobaltinitrite-methemoglobinemia with that generated by sodium nitrite indicates that only 3 of the 6 equivalents of nitrite in each mole of cobaltinitrite are released under biological conditions to react with oxyhemoglobin. A comparison between in vivo and in vitro methemoglobinemias indicates that the major portion (75–80%) of injected sodium nitrite is apparently inactivated by processes other than reaction with oxyhemoglobin; the same is true of cobaltinitrite, but whether in vitro or in vivo only 3 nitrite equivalents per mole are released. Thus, methemoglobin formation by cobaltinitrite accounts for less than half of the cyanide inactivated in pretreated mice. In the case of sodium sulfide, however, the entire protective effect can be accounted for on the basis of methemoglobin formation and subsequent trapping of the hydrosulfide ion. Cobaltinitrite blocks methylene blue stimulation of methemoglobin reductase activity in concentrations that do not produce excessive hemolysis. This effect may be shared by cobaltous chloride, but it is observed only at concentrations that produce gross hemolysis. Indeed, hemolysis produced by 2–8 m m cobaltous chloride would appear to account for the inhibition of methemoglobin reductase activity ascribed to this salt by others.