The purpose of the present paper was to assess the variation in light output from clinical mammographic screens, particularly in relationship to age and radiographic usage. The light output from 156 mammographic cassettes from 17 clinical centres was measured on site using a novel light detector. Survey information on cassette age and usage was collected. Where possible, screen manufacture date was supplied by the vendor. One cassette from each centre was measured at a standard laboratory to allow normalization of all measurements. Three centres were found to have cassettes in clinical use whose light outputs varied by an amount incompatible with current Royal Australasian College of Radiologists (RACR) standards. These centres used cassettes whose ages varied widely. No correlation between light output and radiographic usage was found. Older cassettes were found to have a slight increase in light output. A small reduction of light output was measured between screen regions of high and low radiation exposure histories. Replacement of mammographic screens should not be based on an expectation of light output reduction due to age or usage. Rather, all mammographic screens should be replaced simultaneously as indicated by factors such as failure of screen-film contact or the presence of screen-related image artefacts.