To ascertain in vivo whether recombinant human erythropoietin has a neuroprotective effect on the cortex during subarachnoid hemorrhage, 56 rabbits were divided into the following groups: Group 1 control sham operated plus placebo (n=14; saline solution - NaCl 0.9%); Group 2 control sham operated plus recombinant human erythropoietin (n=14); Group 3 subarachnoid hemorrhage plus placebo (n=14); Group 4 subarachnoid hemorrhage plus recombinant human erythropoietin (n=14; intraperitoneal administration of recombinant human erythropoietin immediately after inducing subarachnoid hemorrhage). In none of the Groups 1 and 2 animals was subarachnoid hemorrhage induced. In Group 3 rabbits, an increase in locomotor activity (open field apparatus) was observed 24, 48 and 72 h after surgery, and the mortality rate was 42.9% within 72 h after surgery, and, no increase in locomotor activity was observed in Group 4 rabbits, which survived for at least 72 h. Our findings suggest that recombinant human erythropoietin may be of benefit in the treatment of subarachnoid hemorrhage.