Contrast sensitivities were measured in human observers of different ages in response to sinusoidal grating patterns generated by laser interferometer and cathode ray tube (c.r.t.). The former method bypasses the effects of the optical media to assess directly the contrast sensitivity of the retina/visual system, while the over-all assessment including optical media is made to the c.r.t. display. The ratio of the two sets of values gives the contrast ratio of the optical media. The latter assessment does not include the effects of wide-angle light scattering which were measured by comparing the responses of eyes with a Perspex lens implant or without any lens at all (i.e. without the major source of scattering) with natural eyes. With increasing age, c.r.t. contrast sensitivities remained steady until the sixth decade when they declined while laser interference fringe contrast sensitivities declined continuously, apart from some abnormally low results for teenage observers. In contrast, the optical contrast did not vary systematically with age nor did its rate of change with spatial frequency vary. The latter was always less than the rate for laser interference fringe contrast sensitivities indicating that the retina/visual system always sets the limit to visual resolution. Wide-angle light scattering was not found to contribute significantly to the changes in ageing nor did reduced retinal illumination, as reproduced by reduced pupil diameter or viewing through a neutral density filter. We therefore conclude that the age-related deterioration is primarily caused by changes within the central nervous system rather than the optical media, the transmission quality of which remains unaffected.