This present study examines the lexical frames in natural disaster news with the goal to create a list of commonly-occurring lexical frames in this news genre as a resource for students and teachers. A lexical frame in this study is defined as a sequence of words with a slot of one word within the frame that can be filled by two or more variants. To investigate this structure, a 188,637-word corpus is constructed using natural disaster news from four online news sources. The period of data collection is one year to include natural disasters that are seasonal. Lexical frames included in the analysis must be five words in length and occur over 40 times per million in at least three different news and in at least three out of the four news subcorpora. In total, 85 lexical frames are identified and they are grouped into three structural categories (verb-based, other-content-word and function-word frames) and four functional categories (stance expressions, discourse organizers, referential expressions and news content lexical frames). The analysis shows that the study of lexical frames in a specialized corpus can be a valuable resource for both students and teachers with regard to learning and instruction of writing in this genre.