During interaction, people coordinate in verbal (e.g., syntactically and semantically) and nonverbal (e.g., gestural and prosodic) ways. This alignment has been suggested to be a result of grounding or priming. In both cases, visual cues assist understanding. This study explores how widely and how much participants align in a text-only environment. Forty-two participants debated a topic via Instant Messenger with a confederate. Using length analyses, LIWC, and LSA, results show punctuation and semantic alignment above chance between interlocutors, and an increase in this alignment over time. Affective alignment and alignment in parts of speech are weak, and the nature of the debate nor nonverbal cues affected alignment. These results extend previous theories of verbal alignment to text- only environments in which interlocutors lack visual cues during interaction and propose theoretical implications for alignment. However, lack of nonverbal alignment departs from face-to-face findings, and theoretical implications for such results are suggested.