A study examined good and poor readers' inferencing strategies as they read unambiguous texts. Subjects, 42 fifth-grade students from 11 classrooms in a Boston metropolitan area urban school district, read 2 well-constructed mystery stories divided into 6 episodes--each episode ended with the introduction of a new clue related to solving the case. Subjects were asked general prompting questions at the end of each section. The verbal reports from each story were combined to form a protocol for each student. Protocols were examined to determine the type of inferencing strategies used. Results indicated that (1) children frequently engaged in a number of inferencing strategies; (2) good and poor readers appeared to use a similar repertoire of inferencing strategies; however, (3) poor readers appeared to accept unconventional interpretations of stories. Findings suggest that teachers may well be advised to emphasize a number of direct instructional activities which help students focus on textual materials rather than emphasizing strategy training. (Three tables of data are included; 24 references are attached.) (RS)