The purpose of this study is to reveal the differences between SMS dialogues and telephone conversations in terms of statistical distribution of lexemes and conversational structures. In order to do so, we have built two corpora, MoCoDa-cqp and Verbmobil-cqp. Based on the SMS dialogues of MoCoDa and telephone conversation data of German Verbmobil Corpus under the topic “Appointment Scheduling”, this study looks into the dialogue structures taking media differences into account, in addition to the statistic distribution of vocabulary and their part of speech. A dialogue structure consists of adjacency pairs of dialogue acts. With the different pairs and their frequency, we have attempted to find the prototypical dialogue structure. As a result, the dialogue structures varied depending on the medium of the conversation. It also became clear, that the topic of the conversation has a great influence on the dialogue structure. Through this study we have observed that the conversation participants used distinctive vocabularies depending on the medium. This result is closely related to the intimacy between the participants and/or the fact that the media environment, communicated in either written or spoken language, has influence on the vocabularies used. For the investigation of the distinctive lexemes of each corpus we adopted the Distinctive Collexeme Analysis framework proposed by Gries and Stefanowitsch(2004).