Introduction <p>The evidence for assessing whether a single stab injury to the trunk was inflicted by another person in an assault or self-inflicted has primarily been derived from autopsy studies. In this study, we investigate whether victim demographics, crime scene circumstances, and the injury characteristics of assault survivals are interchangeable with homicide characteristics, with a specific focus on cases corroborated by perpetrator confession or eyewitnesses.</p> Methods and materials <p>Surviving victims of assaults including single stab injuries to the trunk were identified in a registry including all cases of clinical forensic medicine assessed by the Swedish National Board of Forensic Medicine between the years 2016 and 2021. Characteristics were compared between corroborated and non-corroborated survivors of assault. We also used a reference population of homicide victims with single stab wounds to the trunk.</p> Results <p>Survivors of corroborated (<i>n</i> = 162) and non-corroborated assaults (<i>n</i> = 223) showed largely similar victim demographics and injury characteristics. Compared to fatal cases, survivors more frequently experienced stabs to the abdomen, left axillary region, and the back, and, in fewer cases, to the bony part of the ribcage.</p> Discussion and conclusion <p>Similar findings among corroborated and non-corroborated assault survivors suggest that most non-corroborated cases likely involve victims of actual assaults. According to our results, injury characteristics do not seem to be directly interchangeable between fatal and survived assaults, due to the differences in injury severity, which highlights the need for cautious interpretation of autopsy-based characteristics in living victims.</p>