Popular sentiment toward the teaching of thinking skills and, in particular, higher order skills, is on the rise. Many people are wondering about the best way for teachers to foster the growth of reasoning skills in their classrooms. This study attempts to improve the teaching of reasoning by investigating teaching behaviors or characteristics that might be related to the teachers' reasoning ability. Since there are many science teacher variables that could be related to reasoning ability, a sequential approach was adopted to investigate the problem. Researchers first investigated the degree of clarity of concrete and formal reasoning teachers exhibited in their classroom teaching, their attitudes toward science, and their beliefs about science teaching. Next, researchers replicated their initial findings and investigated other possibilities. The subjects chosen for the study were preservice elementary school teachers. The subjects were in their last quarter of methods instruction before student teaching and were divided into two groups (30 students in the first group, 31 in the second). The critical assessment for this study was formal reasoning ability. The results showed few consistent differences between teachers classified as formal reasoners and those classified as concrete. There was no relationship between reasoning ability and belief in structured science teaching or laboratory oriented science, degree of clarity in the student teaching setting, or learning style. Some speculations about the nature of preservice teachers were advanced. (CW)