From November 1951 to December 1957, all patients with untreated breast cancer admitted to the Radium Centre in Copenhagen were randomized before their operability was evaluated into two groups, if the patients were operable, viz. simple mastectomy with postoperative x-ray treatment or extended radical mastectomy. Twenty-five-year results are presented, showing no difference in survival or recurrence-free survival of the operable patients. Histological grading was performed in nearly all cases. Patients with grade 1 tumours had a better survival than grades 2 and 3, but there was no difference in survival between the two treatment groups, when histological grading was taken into account. Histological node positive patients had more grades 2 and 3, tumours, whereas node negative patients had more grade 1 than grades 2 and 3 tumours. Premenopausal women had a significantly better survival than postmenopausal in all stages.