. The presence of openly non-binary gender individuals on social networks is growing. However, the relationship between gender, activity, and language in online interactions has not been extensively explored. Lack of understanding surrounding this interaction can result in the disparate treatment of non-binary gender individuals on online platforms. We investigate patterns of gender-based behavior identity on Twitter, focusing on gender expression as represented by users’ expression of pronouns from eight different pronoun groups. We find that non-binary gender groups tend to receive substantially less attention in the form of likes and followers compared to binary groups. Additionally, non-binary users send and receive tweets with higher toxicity scores than other groups. This study identifies differences in the language and online activity of users with non-binary gender identity, and highlights a need for further evaluation of potential disparate treatment by algorithms used by online platforms.