The present study was designed to analyze the usefulness of plain radiographs in evaluating bone healing. Rabbit tibiae were osteotomized, externally fixed, and allowed to heal for 3-8 weeks. Bones were harvested, x-rayed, and tested to failure in a dynamic torsion tester. AP and lateral radiographs of 10 rabbit tibia pairs and 10 individual rabbit tibiae were selected randomly for use in a questionnaire, given to 93 physicians who routinely assess fracture healing to evaluate clinicians' ability to assess bone strength. The results indicated that clinicians can differentiate the relative strength of bones by comparing two sets of radiographs. However, the strength determination from a single set of radiographs of a fracture is unreliable, the tendency being to evaluate the fracture to be weaker than it actually is.