Four hundred sixty patients enrolled in a randomized trial of involved field (IF) and extended field (EF) radiotherapy for Hodgkin's disease Stages I and II in the years 1967 to 1973 have been followed to a maximum of 13 years. Minimum time at risk is 6 years, and median follow-up is 8 years. Actuarial survivals are 85% IF and 87% EF at 5 years. The overall standardized risk ratio comparing IF with EF is 1.3, implying a 30% excess mortality in IF, a nonsignificant difference. For males the risk ratio is 1.7 and significant, whereas for females it is 0.6, a nonsignificant reduction in risk with IF therapy. Extension-free survival was significantly better in the EF group than in the IF within 2 years after treatment, and that benefit persists to the current follow-up, with extension-free survivals of 42% IF and 64% EF at 5 years. The risk ratio is 1.7. Favorable survival is significantly correlated with initial characteristics of female sex, age younger than 40 years, and histologic type nodular sclerosis or lymphocyte predominance. Histologic type is the most powerful predictor in the total series, but its prognostic value is not seen in patients staged with laparotomy. Six cases of leukemia have been reported in this series, among whom less than one case would be expected at general population rates. All leukemias have occurred in 167 patients who required chemotherapy for extension of Hodgkin's disease after initial radiotherapy, implying an increased risk following chemotherapy of more than 200-fold.