Abstract: The subject of the research is the multimodal elements in 100 articles of electronic edition of The New York Times of March, 2017. Solntseva examines various kinds of such elements, in particular, photo, video, hyperlink to associated articles, hyperlinks to the front-page news, subscription offer, advertising and technical links. The author focuses on frequency with which these elements can be found in electronic articles. The author also analyzes the dependence of the frequency of multimodal elements on their relevance. The author applies the method of empirical analysis (observation) and describes kinds of multimodal elements in the aforesaid edition using the method of continuous sampling as well as uses the method of classification depending on the frequency of each element in a given or other articles. The main result of the research is the author's description of 5 relevance levels of multimodal elements and discovery of the inverse relation between information relevance and frequency of references to this information in a given or other articles. This dependence can be described using the following formula: the more likely an element is to appear in an article and the more frequent similar elements are found within an article, the lower their relevance for the article is. The author also discovers the inverse dependence between the frequency of a multimodal element and relevance thereof: the more relevant information is, the more frequent this information can be found in articles and the more frequent this information is expressed in the best parts of the article.