This paper presents a summary of the main results of our current research programme on the speech-reception threshold for sentences as a function of the sound-pressure level of interfering noise. The experimental results agree with a simple quantitative model in which every hearing loss for speech is interpreted as the sum of a class-A loss (attenuation), characterized by a reduction in the levels of both speech signal and noise, and a class-D loss (distortion), comparable to a reduction in S/N ratio. The experiments confirm the model's prediction that a hearing aid can compensate for class-A hearing losses but, generally, not for class-D hearing losses. The implies that the problem of having difficulties in understanding speech in a noisy environment, which is the primary handicap of hearing-impaired subjects, is not solved by the hearing aid.