This study investigated the effects of heuristically based assignments and tests on adolescents' written responses to literature. In order to examine both tests and instruction, two separate but related experiments were conducted. Experiment 1 investigated whether the inclusion of guided prewriting on essay tests for two poems and two short stories would result in higher quality written responses than tests that did not include prewriting. Experiment 2 explored the effects of a series of five heuristically based lessons on the posttest essay scores on two poems and two short stories. Approximately 100 11th and 12th graders participated in each experiment. Analyses of the data indicated a significant treatment effect on one of the four test measures in Experiment 1, with students who received the experimental test format performing significantly better than students who received the National Assessment of Educational Progress (NAEP) format test. There were no significant effects for the instructional treatment in experiment 2. Reading ability and genre had significant effects on the quality of students' written responses to literature in both experiments, with higher ability students receiving significantly superior mean scores and all students performing better on tests for short stories than on tests for poems. Implications for teaching and further research into longitudinal training in heuristic procedures are discussed. A seven-page bibliography, a sample NAEP test, a sample experimental test, and a sample lesson follow the study. (Author/JAZ)