The study presents the results of a standardized content analysis comparing the use of Twitter as an information source on the websites of five news media types (quality newspapers, a tabloid newspaper, weekly magazines, broadcasters, and internet only). The theoretical assumption behind the study is that the adoption of Twitter as a source follows the established practices of a particular media type. All articles published online by ten German news media in a period of one month were collected (n = 21,823). A subset of articles containing Twitter-related keywords (n = 496) and 375 tweets cited in these articles was explored in detail. Our focal analysis of n = 287 articles using Twitter as an information source revealed systematic differences between both media types and article topics regarding the number and style of Twitter references as well as the types of accounts cited. We found that the tabloid paper incorporated the highest number of tweets per article, incorporated tweets primarily in articles on human interest and gossip and cited non-elite sources more often than other media types. Quality papers, weekly magazines, and broadcasters used tweets as sources primarily in articles on politics and economy and cited public actors, such as politicians or officials, more frequently.