The article offers an empirical analysis of the disrespectful online comments—in total, 17,581,659—in Hungary between 2017 and 2019. Considering the name-calling and obscene and abusive phrases as communication practice, we rely on computational linguistics to investigate incivility in comments posted on online news stories and news portals’ social media sites across time, platforms, and topics. The results of the dictionary-based quantitative content analysis show that incivility’s frequency is steady in the examined three-year period. The topic and platform make the real difference: Politics attract more uncivil comments than business news, and less incivility is detected on Facebook than on the news sites. Contrary to popular perceptions, we find that incivility has not become significantly intensive in election campaign periods. Incivility, however, correlates to the dynamics of interaction. The comment-level analysis suggests that uncivil messages are likely to be surrounded by the same communication feature.