An adaptation of the sensorineural acuity level procedure was employed to obtain thresholds under bone- (BC) vs. air-conducted (AC) white noise masking. For the BC masking condition, a Radioear B71 was placed on one mastoid. An Etymotic ER3 with a foam tip placed in the ear on the same side was used to deliver the pure-tone probe stimulus. This transducer was chosen to approximate the Telephonics TDH-39 earphone response characteristic while reducing occlusion effect. For the AC masking condition, the masker and probe were mixed electrically and delivered to the earphone. Masked threshold data, transformed into frequency response curves, demonstrated greater variance of BC vs. AC response across frequency but less high-frequency roll-off than expected from coupler measurements obtained using an artificial mastoid and 6-cm3 cavity, respectively.