Mortality rates of mice and their levels of anti-venom and anti-F(ab')2 antibodies were assessed after three episodes of subcutaneous envenomations with or without treatment with horse F(ab')2. Soluble venom from the Mexican scorpion Centruroides limpidus limpidus was used for these experiments. Repetition of episodes did not induce different mortality rates in untreated mice. F(ab')2 rescued about 85% of the mice in the first two episodes and 66% in the third, without distinction of gender or ostensible side-effects: a suggestion of selection of the most resistant mice. Surviving mice produced in vitro neutralizing antibodies to the scorpion venom and also antibodies to F(ab')2, when injected alone but more so if combined: a possible immunological adjuvant or alarm effect of the venom or of the cascading physiopathology of envenomation. In the few surviving mice, both anti-venom and anti-F(ab')2 antibodies increased significantly after the first envenomation but not thereafter, showing no correlation with mortality rates: a suggestion of their clinical irrelevance, the few hard-to kill mice appeared to resist envenomation by mechanisms other than antibody response. Injection of F(ab')2 alone induced production of detectable anti-venom antibodies in a few mice and injection of venom alone induced that of anti-F(ab')2 antibodies, perhaps due to trace amounts of venom in the high affinity fraction of F(ab')2 and to anti-idiotypic antibodies or polyclonal activity in the envenomation episode, respectively.