Background and objectivesAn attention bias toward hostile stimuli is associated with an increased risk of aggressive behavior. Thus, reducing hostile attention bias may help to reduce aggression in young offenders. We evaluated the remediating effects of Attention Bias Modification (ABM) on hostile attention bias and aggressive behavior in Chinese male young offenders with antisocial tendencies. MethodsInstitutionalized male young offenders (ages 16–18) were recruited and randomly assigned to ABM (n = 28), placebo (n = 28), or waiting list (n = 28). The ABM group received four weeks of training using visual search of emotional faces; the placebo group underwent similar training using visual search of neutral objects. Before and after treatment, aggressive behavior, attention bias toward positive stimuli and hostile stimuli were assessed. ResultsLinear mixed models and hierarchical multiple regression analyses showed that the intervention effect of ABM was moderated by participants’ initial attention bias. Young offenders with a high level of initial hostile attention bias, or with a low level of initial positive attention bias, benefitted most from ABM. LimitationsThe small sample size impedes investigation of the mechanism of the intervention's effects. Further research can determine whether the current results can be extrapolated to the population of female young offenders. ConclusionThe results of the present study suggest that ABM is helpful for young offenders who show the greatest bias toward hostile stimuli and away from positive stimuli. These findings highlight the importance of matching the intervention and the participant. Trial registrationosf.io/vj5rk;