A study assessed the ability of 20 eleventh grade students to recognize particular stylistic features when reading and then to use such features in their writing. Half the subjects were given a model paragraph from an Edgar Allen Poe story, without having the story identified for them, while the other half received only a verbal summary of the story, with the author and genre identified. All students were given a written summary of the paragraphs that followed the model paragraph. These summary paragraphs had been modified from the original in four characteristics (the long sentences in the original were shortened, the tense was changed, all subordination was removed, and the number of adjectival phrases was reduced). The students who had read the model paragraph were instructed to rewrite the summary paragraphs so that they would seem to belong with the model text. The other students who did not see the text were instructed to rewrite the summaries to create a "horrifying" story. All students were given six phrases taken from the original Poe paragraphs that they were supposed to work into their compositions. Between the two groups there was no significant difference in the ability to create a horrifying story. There were also no significant differences between the groups in their use of three of the four modified characteristics: length of sentence, and use of subordination and adjectival phrases. Thus, except for tense, a model text did not help the students to write in the horror genre. (HTH)