Purpose <p>This study examines the differential effectiveness of communication skills interventions for situationally and characterologically violent male-perpetrated intimate partner violence (IPV).</p> Method <p>Using data from Babcock et al.'s (2011) proximal change experiment, latent class analysis was used to categorize IPV perpetrators into situational or characterological groups based on patterns of physical aggression, psychological abuse, and coercive control. Differential effectiveness of two communication-based interventions, "Accepting Influence," and "Editing out the Negative," and a placebo control condition was tested between the two groups.</p> Results <p>We found that men in the situationally violent group reported significantly higher levels of increases in positive affect after both interventions compared to men in the characterologically violent group. On psychophysiological outcomes, significant interactions emerged showing that both active&#xa0;interventions led to a greater increase in post-intervention skin conductance for situationally violent men than for characterologically violent men. However, there was no evidence for differential responding based on other physiological variables, observed behaviors, or partners’ self-report.</p> Conclusions <p>The current study contributes to scalable and practical solutions for treatment matching in experimental designs for IPV.</p>