Background: A neurobiological perspective has become accepted as a valuable approach for understanding anti-social behaviour. There is literature to suggest that, in non-offending populations, psychological treatments affect both neurobiological measures and clinical presentation. A theoretical position to this effect has been adopted with respect to offender treatment, but there has been no systematic review of empirical literature on this point. Aims: This study aimed to ascertain from published literature firstly whether there is evidence of change in neuropsychological or physiological measures after behavioural treatments/programmes for people with anti-social behaviour and secondly whether these neurobiological changes are associated with behavioural change. Method: A systematic search strategy was formulated to include studies considering 'neurobiological factors', 'anti-social population', 'treatment' and 'treatment outcome'. The Maryland Scientific Methods Scale was used to select relevant studies of sufficient methodological quality. Results: Eleven studies were found, only one with adults. Overall, the values of specific neurobiological risk factors, particularly of basal cortisol, become less abnormal following intervention. There was some evidence for a link between change in neurobiological functioning and behavioural improvement. Conclusions: Findings, although provisional, may provide new insights into the underlying mechanisms of interventions for anti-social behaviour. Future studies that include pre-treatment neurobiological assessment could help reveal physical vulnerabilities that interventions should target to improve treatment efficacy, and provide for objective, independent corroboration of change.