This study was designed to research whether the conditions that give rise to exemplar effects in experimental designs are present in the real world, specifically by conducting a comprehensive content analysis of news articles in weekly U.S. news magazines. Exemplification studies the relationship between examples and the larger population they represent, and how examples effect consumer's perceptions and behaviors (Zillmann & Brosius, 2000). In experimental design several independent variable have been tested and have shown that people's perceptions fall largely in line with the emphasis of the exemplars presented. A stratified random sample of magazines, representative of a whole year, was obtained for TIME and Newsweek. An intercoder reliability test was performed with 11% of the sample. Eighty-seven articles met the coding requirements and generated 873 exemplars. This research developed a significant number of operational definitions and procedures for content analysis of exemplars. A discussion of issues arising in of content analysis that were not manifest in experimental designs is presented such as non-news articles, the presence of bias, and multiple article foci. The concept of primary base rate data, the reasonable reader test, and expanded definitions of visual exemplars are also presented.Several of the conditions that gave rise to exemplification effects in experimental designs were present. Eighty percent of articles had more exemplars than counterexemplars; Sixteen percent of articles contained perceptually enhanced base rate data; Ninety percent of articles contained no ratio data—meaning a judgment of how representative the exemplars were was not possible. The remaining 10% were considered to be non-representative. Some elements considered to give exemplars more influence were not common in weekly U.S. news magazine articles. Direct quotes were used in only 27% of exemplars, with anecdotes comprising 51%. Similarly, the majority of exemplars (52%) came from non-attributed sources or official reports. Vivid emotion was present in only 2% of exemplars. In addition, 31% of articles were judged to be about a single exemplar, with no counterexemplars present. The most common type of image used were innocuous, with threatening images used the least. Fifty-six percent of exemplar sources were not attributed to a gender, 33% of exemplar sources were male and 7% were female. Similarly, 54% of exemplar subjects did not specify a gender, while 25% were about males and 6% were about females.