Building on recent research that challenges the notion that norm violations in online discussions are inherently detrimental, this study relies on a distinction between incivil and intolerant user comments and investigates how online users perceive and react to these distinct forms of antinormative discourse online. Conducting a preregistered factorial survey experiment with a nationally representative sample of n = 964 German online users, we presented participants with manipulated user comments that included statements associated with incivil (profanity; attacks toward arguments) and intolerant discourse (offensive stereotyping; violent threats). The results show that intolerant statements consistently lead to higher perceptions of offensiveness and harm to society as well as an increased intention to delete the comment containing the statement, whereas incivil statements do not. An exploratory multiverse analysis further suggests that these effects remain robust across a variety of analytical decisions. Lay Summary Online discussions often violate social norms. People can be rude, use offensive language, or even make threats. But research shows that not all norm violations are equal. Incivility (e.g., using profanity) is primarily a way of getting attention and can still allow for meaningful discussions. Intolerance (e.g., using violent threats), on the other hand, is meant to cause harm and goes against what is accepted in a democratic society. Our study asks whether online users react differently to incivil and intolerant statements in user comments. The results show that users find intolerant comments worse and consider them more harmful than incivil comments. These findings are consistent even when looking at the data in different ways.