Computer-mediated communication (CMC) is a relatively new field made even more difficult to investigate by the continuous progress and changes of the digital world. There is a multitude of research on CMC from a linguistic perspective; however, the vastness of the online world makes it difficult to determine to what extent it is a welcoming or hostile environment. The existing studies of polite and impolite language use in CMC provide some insight into the type of environment found online though there is a notable lack of information regarding the overall proportionality of impolite language use, specifically violent and aggressive language use. This study aims to explore to what extent violent and aggressive language use is present in user generated comments found on the social media platform YouTube by investigating the semantic fields of violence and aggression. It is a corpus-based study of the topmost comments collected from the most popular videos in the year 2023 in the YouTube category of Reviews. The quantitative content analysis reveals that both violent and aggressive language use is present in the comments though infrequently with violent language use having a normalized frequency of 2.1 times per thousand words and aggressive language use 2.6 times per thousand words. The qualitative pragmatic analysis of the observed instances shows that only one instance carries out the illocutionary act of violence in addition to the locutionary act. Based on these findings it appears violent and aggressive language use is relatively infrequent concerning more neutral topics and in most cases the communicators do not have violent or aggressive intent in their messages.