This paper reviews the epidemiology of lung cancer in uranium miners in northern Ontario whose cumulative exposure was relatively low and who were exposed only for a short period of time. The development of the "epidemic" is demonstrated chronologically and in terms of latency. An apparent difference between the effects of short-term and of more prolonged radiation is shown. A dose-response effect is demonstrated at all levels. Analysis of tumor type suggests that there may be a difference in dose-response between the two principal types encountered. The relationship of the geographic location of exposure to the final residence at death is reviewed.