The present study has as main aim to compare the frequency of physical violence and injuries among couples that report bidirectional and unidirectional violence in a community sample in Arequipa, Peru. This comparison also accounted for the level of severity in violence and injuries. Several previous studies have suggested that reciprocal violence could be the most common pattern of violence in intimate relationships. Relationships, where both members are violent to each other, could also be at a higher risk of violence, both in frequency and severity, since symmetrical interactions can provoke it to escalate. Likewise, injuries have been found to be more common in bidirectional violence. However, this type of violence has not been previously studied in Peru. Additionally, a comparison between genders of the participants was made. Different studies, including one in Peru, have suggested that men and women have the same risk of suffering violence from an intimate partner, so this objective was directed to help clarify this, at least in a community sample. The participants were 280 people with ages between 18 and 65 years, 50.7 % women and 49.3 % men, who were married or cohabiting and living in three districts of Arequipa. The Conflicts Tactics Scale Revised was used, in the version that assesses behaviors of both members of a couple (Straus et al., 1996). The subscales of physical violence and injuries, and the levels proposed by the scale of minor and severe, where used for the analysis. According to the results, 73.5 % of the participants reported at least one incident of violence, 61.4 % presenting bidirectional violence and 12.1 % unidirectional violence. The information provided by these participants was the one used for the analysis. In addition, a higher number of participants, of the ones in the unidirectional violence group, reported perpetrating violence than participants that reported being victims of violence by their partner. In comparisons across gender, no significant difference in the frequency of violence or injuries was found, neither in relationships with bidirectional nor unidirectional violence nor in minor or severe violence or injuries. In the comparison across types of violence, significant differences between unidirectional and bidirectional violence were found in both levels of severity of violence (U = 625.0, p < .001, for minor violence; U = 1335.0, p < .001, for severe violence), where the effect size for minor violence was moderate (rrb =.506, p < .001) and for severe violence was weak (rrb = .301, p < .001). Differences were also found in frequency of injuries between these two types of violence (U = 1898.5, p < .001, for minor injuries; U = 2225.0, p < .01, for severe injuries). In this case the effect size was weak for both comparisons (rrb =.255, p < .001, for minor injuries; rrb = .209, p < .01, for severe injuries). The results show that men and women have a similar risk of suffering violence and having injuries because of it. However, they also describe that a higher risk exists in couples where violence is bidirectional, and that should be taken into account in intervention aspects on the subject. For future research on the topic, other types of intimate partner violence should be considered, such as psychological or sexual abuse, as well as other variables that could affect the experience of the victim, such as controlling behaviors. Furthermore, research that clarifies interactions that lead to violence by members of the couple and aspects that lead to desistance in the use of violence is needed. [ABSTRACT FROM AUTHOR]