Spread of excitation in the cochlea places fundamental limits on the interpretation of audiometric pure-tone hearing loss as a simple map of dysfunction along the cochlear partition, and histologic evidence from human temporal bones will be presented to demonstrate the insensitivity of the audiogram to variations in pathology in the case of low-frequency hearing loss. This article will describe a clinical procedure using simultaneous pure-tone masking to improve upon the localization of cochlear disease, particularly for low-frequency hearing losses, and a model for using the Articulation Index (AI) to develop prognoses for hearing aid performance in these cases, which can then be tested. Fourteen patients with low-frequency hearing loss were divided into two groups based upon threshold shifts caused by a pure-tone masker: those that showed normal low-frequency threshold shifts and those that showed marked shifts at frequencies below the masker, indicating greater loss of function than shown by the unmasked audiogram. Hypothetical audiograms were then generated to model a complete loss of apical function for all patients. Measured speech recognition scores were then compared to AI predictions for the actual and hypothetical audiograms. Best agreement for the patients showing normal masking shifts was between the measured scores and the AI for the actual audiogram, whereas the best agreement for the patients showing marked shifts was with the AI for the hypothetical audiogram. The implications for hearing aid recommendation and fitting in these cases are discussed.