Accurate communication of the severity of violence in intimate relations is essential for the appropriate evaluation of offenders and victims in contexts such as court trials, custody cases, and the continuation of relationships. Using a new paradigm, this study quantifies discrepancies in how the severity of violence is communicated in texts written by offenders, victims, and bystanders who witness violence. The study was conducted in two phases, where participants were randomly sampled from the same population to participate in either Phase 1 or Phase 2. In the first Phase, witnesses (narrators) provided nine narratives about self-experienced intimate partner violence and rated the violence’s severity; then in the second Phase non-witnesses (recipients) read all the narratives and rated the severity of the violence. Four types of perceptual differences (calibration, accuracy, gender, and role perceptual differences) were identified when rating the severity of three types of violence (psychological, physical, and sexual) as communicated by three types of witnesses (victims, offenders, and bystanders) of violence in heterosexual, romantic relationships. Several novel findings were made related to a strong perceptual difference in calibration, i.e., a tendency for the recipient to rate the violence more severely than the narrator, where this effect was mainly found for victims and bystanders, but not for offenders. Also, the calibration effect was largely seen in the sexual and physical, but not psychological, narratives. The recipients’ accuracy was considerably lower for psychological rather than sexual violence. Finally, the validity of the method was confirmed by replicating earlier findings on perceptual differences in roles where witnesses rated violence more severely than victims or offenders and women were rated more severely than men, which was especially true for male raters. These results suggest systematic perceptual differences in severity ratings and may have substantial implications for victims and offenders in real-life settings. These findings may potentially be used to ameliorate the negative effects of perceptual differences.