Recent scientific developments in auditory psychophysics and cochlear physiology offer new insights into the nature of both normal and impaired hearing. However, it is not always clear how these insights can be put to practical use in a clinical context; for example, in the design and fitting of hearing aids. While a diagnosis in terms of cochlear pathology might appear to be desirable, it cannot, at present, be used in the fitting process. As a consequence, the time needed to acquire the additional information for a pathological diagnosis cannot be clinically justified. Computer models of the hearing of an individual patient are proposed as tools to bridge the gap between laboratory science and audiological practice.  In this approach, measurements of critical qualities of a patient hearing (thresholds, tuning and compression) are used to shape a computer model that can be used to explore the benefits of alternative prostheses in the absence of the patient. These models are called ‘hearing dummies’ by analogy with the use of a tailor’s dummy in dressmaking. Patients with sensory-neural diagnoses and similar audiograms can have very different tuning and compression characteristics. This suggests different underlying pathologies that indicate different prosthetic strategies particularly when seeking to improve the patient’s capacity to follow conversations in noisy environments. This talk will illustrate the wide range of different patient hearing profiles found in our laboratory and show how they can be captured in physiologically-based computer models. These models can then be used to predict the response of the listener to normal acoustic events with and without prostheses.