Previous studies using bolus intravenous injections of sodium cyanide have been used to model the sudden exposure to high concentrations of cyanide that could occur on the battlefield. This study was designed to develop a model that would simulate the type of exposure to cyanide gas that could happen during actual low-level continuous types of exposure and then compare it with the bolus model. Cardiovascular and respiratory recordings taken from anesthetized dogs have been used previously to characterize the lethal effects of cyanide. The intravenous, bolus injection of 2.5 mg/kg sodium cyanide provides a model in which a greater than lethal concentration is attained. In contrast, our model uses a slow, intravenous infusion of cyanide to titrate each animal to its own inherent end point, which coincides with the amount of cyanide needed to induce death through respiratory arrest. In this model, therapeutic intervention can be used to restore respiration and allow for the complete recovery of the animals. After recovery, the same animal can be given a second infusion of cyanide, followed again by treatment and recovery, providing a reproducible end point. This end point can then be expressed as the total amount of cyanide per body weight (mg/kg) required to kill. In this study, the average dose of sodium cyanide among 12 animals was 1.21 mg/kg, which is approximately half the cyanide used in the bolus model. Thus, titration to respiratory arrest followed by resuscitation provides a repetitive-use animal model that can be used to test the efficacy of various forms of pretreatment and/or therapy without the loss of a single animal.