The process of collecting text-messages for scholarly investigation poses researchers with a number of challenges related to communicative context, privacy, and data transferal. This paper explains a data collection method that largely overcomes these challenges. The method is a form of Discourse Completion Task in which study participants are asked to send from their mobile phones the text-messages they would send in certain hypothetical scenarios. While similar methods have been used in some previous studies, this paper presents the first full explanation and justification of the text-message DCT method, and includes an explanation of the how the method was used successfully in a study of punctuation marks in text-messages.