Bone-conduction thresholds were obtained at the mastoid process and frontal bone on normal listeners and on cases with surgically confirmed middle ear lesions. The results suggest that serious consideration be given to the use of unoccluded frontal bone-conduction measurements as a clinical testing procedure, especially in diagnostically challenging cases with mixed loss. It must be understood that the appropriate calibration procedures be utilized for such tests. It is further suggested that measurements of absolute and relative acoustic impedance and tests for the occlusion effect be employed in conjunction with the measurements at the mastoid and frontal bone when the conductive component is small and where the question of surgical exploration demands extensive audiological information.