The hypothesis that circulating cortisol levels influence the outcome of social interactions in rainbow trout was tested. Juvenile rainbow trout Oncorhynchus mykiss were given a single intraperitoneal (i.p.)implant containing either cortisol (110 mg kg–1 fish), or cortisol plus the glucocorticoid receptor antagonist RU486 (mifepristone; 1100 mg kg–1 fish), and sampled after 5 days of social interactions with either a similar sized (<1.5% difference in fork length)or smaller conspecific (>5% difference). Within size-matched pairs of fish,cortisol treatment significantly increased the probability that the treated fish within each pair became subordinate, an effect that was abolished by simultaneous administration of RU486. Cortisol treatment also reduced the usual success of the larger fish within a pair to preferentially become dominant from 86% to 40% of pairs. To investigate one potential mechanism underlying the apparent effect of cortisol in predisposing trout to low social status, fish were treated with cortisol or cortisol+RU486 for 5 days, after which brain monoamines [5-hydroxytryptamine (5-HT); dopamine (DA)] and their major metabolites [5-hydroxyindolacetic acid (5-HIAA);3,4-dihydroxy-phenylacetic acid (DOPAC)] were measured. Significant increases of serotonergic activity ([5-HIAA]/[5-HT] ratio) were detected in the telencephalon with cortisol treatment, an effect that was eliminated by simultaneous administration of RU486. Also, cortisol treatment significantly decreased dopaminergic activity in the telencephalon. Somewhat surprisingly,the effects of cortisol treatment on monoaminergic activity in the hypothalamus were opposite to those in the telencephalon. Moreover, in no case did administration of RU486 abolish these effects. These results suggest that the effects of cortisol on social status in rainbow trout may be mediated via the modulation of central signaling systems and subsequent changes in behaviour and/or competitive ability, although the exact site of action in the brain remains uncertain.