A large group of consecutive patients were studied who had been treated for a primary breast carcinoma by a radical mastectomy at the Netherlands Cancer Institute, during the period of 1960-1974. One of the objectives of that retrospective study was to record the results of treatment of a historical group of patients in such a way that they could presumably serve as a reference for new methods of treatment of breast cancer. To verify the validity of this presumption, the present study compares the treatment results of patients from the EORTC trial 10801, which addresses the value of breast conserving procedures, with that of matched controls from the historical group mentioned above. Matching was carried out with respect to the following prognostic factors: age, tumor size, localization, (number of) positive axillary nodes and grade of malignancy. Analysis of the results strongly suggests that the prognosis for the patients from the historical group is worse than for both groups of patients from the trial 10801. Therefore, this study again confirms that a historical group is not suitable as comparative material for new ways of treatment of breast cancer, and that prospective randomized studies are required to test new treatment schedules.