After mastectomy, the combination of Grade III tumour with nodal involvement incurs a high risk of local and regional relapse. We have previously reported (ISRO Meeting, Kyoto 1992) initial findings from a trial in randomly allocating such patients to receive postoperative RT (45 Gy in 15 fractions over 3 weeks), or not. From 9/85 to 9/91, 77 patients were recruited. All patients now have a minimum of four years from randomisation.The results are:RT group (n=36)No-RT group (n=41)“p” valueLocoregional récurrence5 years21%64%<0.0001Free of any récurrence5 years41%21%0.06Survival5 years56%42%0.23The benefit of RT in controlling locoregional disease is re-emphasised by these mature data. The trial was closed after a short period of recruitment, having achieved significance for this end-point. The possibility that RT might have a wider role—in prolonging overall disease-free interval (and possibly survival)—now emerges, but cannot be unequivocally established from this study.