Laboratory experiments on pairs of male male of the genus Xiphophorus were carried out to test the hypothesis that the duration of a pre-experimental isolation period would modify the effects of prior residence on aggressive behaviour and dominance. Resident and intruder were isolated either both for 3 or both for 22 h (Exp. 1), or one for 3 and other other for 22 h (Exp. 2). Prior residence effects were only evident when the resident had been isolated for 3 h. Other connections between isolation conditions, reciprocal aggression and final dominance were also analysed, as was the influence of size differences between the opponents.