To examine the effects of inductive and programmed instruction in the close reading of poetry, four control classes of urban 11th graders receiving inductive instruction in poetry were compared with four equivalent experimental classes using programmed textbooks. Both groups were given free response tests on two poems: one poem, included in classroom instruction, was presented on pre- and post-tests, while a second poem, not included in the instruction, was presented only in the post-test. Content analysis, using "Elements of Writing about a Literary Work" by Purves and Rippere, revealed significant post-instruction differences in favor of the control group for perception, interpretation, and fluency categories, but no differences for the evaluation or engagement categories. It could not be shown that the programmed group altered either the quantity of responses or its fundamental approach. Results indicated that a more thorough evaluation of programmed literature texts should be undertaken by publishers before curriculum makers and teachers employ these materials in their classrooms, and that inductive teaching is a promising method of teaching poetry. (Author/MF)