All underground mines have higher radon levels than are found in surface air. Ventilation is the primary method of controlling radon levels. Fourteen cohort and seven case-control studies done on underground miners are reviewed; they include many types of ore. Only five of the studies deal with more than 100 lung cancer deaths. Variations in the attributable risk are given. Some generalizations can be drawn from these studies: the longer the follow-up, the greater is the attributable risk, even though the relative risk is reasonably constant. The induction-latent period is quite variable but is shortened by high exposure rates, by cigarette smoking, and by increasing age at start of mining. The predominant histological type of lung cancer among miners changed from small-cell undifferentiated for short follow-up time to epidermoid after long follow-up times. With short follow-up time, a multiplicative interaction between smoking and radiation was indicated, but, with long follow-up time, the two factors appear to be simply additive. This difference is probably due to the shortened latent period among cigarette smokers, not to synergism.