Hereditary deafness has proved to be extremely heterogeneous genetically with more than 40 genes mapped or cloned for non-syndromic dominant deafness and 30 for autosomal recessive non-syndromic deafness. In spite of significant advances in the understanding of the molecular basis of hearing loss, identifying the precise genetic cause in an individual remains difficult. Consequently, it is important to exclude syndromic causes of deafness by clinical and special investigation and to use all available phenotypic clues for diagnosis. A clinical approach to the aetiological investigation of individuals with hearing loss is suggested, which includes ophthalmology review, renal ultrasound scan and neuro-imaging of petrous temporal bone. Molecular screening of the GJB2 (Connexin 26) gene should be undertaken in all cases of non-syndromic deafness where the cause cannot be identified, since it is a common cause of recessive hearing impairment, the screening is straightforward, and the phenotype unremarkable. By the same token, mitochondrial inheritance of hearing loss should be considered in all multigeneration families, particularly if there is a history of exposure to aminoglycoside antibiotics, since genetic testing of specific mitochondrial genes is technically feasible. Most forms of non-syndromic autosomal recessive hearing impairment cause a prelingual hearing loss, which is generally severe to profound and not associated with abnormal radiology. Exceptions to this include DFNB2 (MYO7A), DFNB8/10 (TMPRSS3) and DFNB16 (STRC) where age of onset may sometimes be later on in childhood, DFNB4 (SLC26A4) where there may be dilated vestibular aqueducts and endolymphatic sacs, and DFNB9 (OTOF) where there may also be an associated auditory neuropathy. Unusual phenotypes in autosomal dominant forms of deafness, include low frequency hearing loss in DFNA1 (HDIA1) and DFNA6/14/38 (WFS1), mid-frequency hearing loss in DFNA8/12 (TECTA), DFNA13 (COL11A2) and vestibular symptoms and signs in DFNA9 (COCH) and sometimes in DFNA11 (MYO7A). Continued clinical evaluation of types and course of hearing loss and correlation with genotype is important for the intelligent application of molecular testing in the next few years.