AbstractIsolation‐induced intermale fighting in laboratory mice can be dramatically reduced under most circumstances by castration. This behavior in castrates may, however, be restored, or even accentuated, by androgen replacement. Experiments on the effects of sex steroids on such fighting in castrated mice, which, for want of a better term, are designated as “aggressive,” have been recently described. These mice are housed with a female until 10 days after siring a litter and are, thereafter, housed individually for a further 14 days before castration and subsequent hormone treatment. Such mice show substantial levels of fighting in “standard‐opponent” tests even before isolation. Although castration results in reduced fighting in these mice, this behavior is rarely completely abolished in all individuals. It seems likely that steroid treatment of aggressive mice maintains or intensifies an already present motivation. Treatments in these studies consisted of daily oil‐based intramuscular injections for 14 days preceding and throughout behavioral testing. The standard‐opponent tests were 7 min encounters with adult, subordinate, grouped males in the cleaned home cages of experimental mice. The steroids investigated included estradiol benzoate (EB), 19‐hydroxytestosterone (19‐OHT), androstenedione (A), testosterone (T), and Sα‐dihydrotestosterone (DHT), either singly or in combination. The results suggest that (a) on a dosage basis, estrogens were at least as effective as androgens in maintaining fighting in castrated aggressive mice; (b) 19‐OHT (one of the metabolic intermediates between testosterone and 17 β‐estradiol) was also effective but somewhat less so than the same dose of EB; (c) the three naturally occurring androgens investigated all effectively maintained fighting at comparatively low doses (50 μg/day) which compares with a replacement dose of 500 μg/day of T in some studies in traditional castrated mice (e.g., Luttge and Hall, 1973); (d) aromatization is not essential for a behavioral action of androgens as DHT, a nonaromatizable androgen, maintained fighting in these mice; (e) whereas a two‐site (central motivational and peripheral penile) action seems probable in the influence of androgens on sexual behavior in castrated rats (e.g., Parrott, 1975), DHT did not augment the action of EB on fighting in castrated aggressive mice, indicating that only a central action of steroid