The effects of sexual variables on shock-elicited aggression (SEA) in rats were studied in four experiments. The results showed that (a) Sprague-Dawley males had higher SEA frequencies than females, but this difference was not statistically significant; (b) Long-Evans males exhibited significantly higher SEA frequencies than females; (c) sham-operated Sprague-Dawley males showed significantly higher SEA frequencies than castrated males, but this difference was diminished by testosterone replacement therapy of the castrate group; and (d) SEA was unrelated to free-field or stabilometer activity measurements. Considerable aggression occurred in all animals, suggesting that although sexual variables affect SEA, they do not play so important a role in SEA as in other kinds of aggression.