The efficacy of corticosteroid therapy in the treatment of shock remains controversial. In order to evaluate this question, the following controlled experimental study was undertaken. There were 44 puppies (2-6 kg) used to examine the effects of methylprednisolone (30 mg/kg) in both hemorrhagic and live Escherichia coli septic shock. In order to isolate the effects of steroid treatment, no volume or antibiotic therapy was given. Arterial, venous, and pulmonary artery catheterization allowed continuous hemodynamic and metabolic monitoring. One control group received steroid treatment and was not subjected to shock. Septic shock was achieved by a rapid bolus infusion of 10(9) live E. coli organisms. Hemorrhagic shock was produced by bleeding the puppy an average of 43% of his blood volume. Four septic and four hemorrhagic shock groups received either no treatment, steroids at the time of shock, or steroids 30 min before or after shock. Cardiac outputs of less than 50% of control values and significant lactic acidosis were documented in all of the shock animals. The septic groups showed more profound alterations in these parameters and a decreased overall survival. No statistically significant differences could be found, however, in the hemodynamic, metabolic or survival figures among the different septic shock groups, or among the various hemorrhagic shock animals. The anticipated preservation of cardiac output and decrease in leakage of lysosomal acid phosphatase were not seen with any treatment schedule. The theoretical benefits of corticosteroid treatment in shock could not be documented in these two models of severe septic and hemorrhagic shock in puppies.