There is considerable interest in offering insight into the mechanisms that might explain why certain perpetrators of intimate partner violence against women (IPVAW) drop out of interventions. Although several socio-demographic variables and attitudes towards IPVAW have been proposed as risk factors for IPVAW perpetrators' dropout, less attention has been paid to alcohol misuse, and its interactions with empathic and cognitive deficits, in the discontinuation of the treatment. Therefore, the main aim of this study was to compare the profile of a carefully selected sample of IPVAW perpetrators, divided into four groups: those who completed the intervention with low (n = 267) and high alcohol consumption (n = 67); and those who dropped out before the intervention ended with low (n = 62).and high alcohol consumption (n = 27). Furthermore, we also studied the initial risk before the intervention started and the o cial IPVAW recidivism during the first year after the intervention ended. Our results revealed that IPVAW perpetrators, especially those who did not complete the intervention and had high alcohol consumption/alcohol misuse, were less accurate in decoding emotional facial signals and presented more errors and perseverative errors than those who completed the intervention. These differences were particularly marked in comparison with those who showed less alcohol consumption. Furthermore, the first group also presented the highest risk (assessed by therapists) and official recidivism rate (o cial records). Conversely, the lowest rate of IPVAW reoffending was presented by IPVAW treatment completers with low alcohol consumption. Hence, our study identifies different targets that should be addressed during the initial stages of interventions in order to prevent or reduce IPVAW dropout, which in turn might reduce violence recidivism in the long term through their effects on emotional information processing and behavioural regulation.