To evaluate the incidence and prevalence of various signs of late morbidity, their time of appearance and pattern of progression during an observation period up to 34 years in breast cancer patients treated with postoperative radiation therapy after radical mastectomy. A group of 71 breast cancer patients received in 1963-1965 aggressive postoperative telecobalt therapy to the parasternal, axillary, and supraclavicular lymph node regions after total mastectomy and axillary clearance. None of the patients received chemotherapy either prior to, or after the irradiation as part of their primary treatment. The prescribed dose to the three lymph node regions was 44 Gy in 11 fractions. Only two of the three fields were treated per day. This total dose was given in 16-17 fractions over 3-4 weeks. Because of the overlap of the supraclavicular and axillary fields, the dose received by the brachial plexus was not the dose that was prescribed. A retrospective dose calculation showed that the total dose to the brachial plexus was 57 Gy, delivered as a complex combination of 1.8 Gy, 3.4 Gy, and 5.2 Gy fractions. This cohort of patients has now been followed to 34 years and the late side effects of the treatment evaluated and scored. This series is unique in the literature. There is no comparable report of a detailed long-term follow-up in a homogeneously treated group of patients with such a high survival, especially among the younger women, where it is almost 50% at 30 years. This is the reason that they were able to develop some of the very slowly evolving injuries. There was progression of many of the late effects in the period between 5 and 34 years. The more serious morbidities have increased progressively over the whole 34-year follow-up period. Ninety-two percent of the long-term survivors have paralysis of their arm. Other neurological findings included unilateral vocal cord paralysis among 5% of the patients, who developed the disease after a median time of 19 years. All of them were left-sided, indicating a mediastinal involvement of the recurrent nerve. Local recurrence or the appearance of a new primary tumor infiltrating or causing pressure on the recurrent nerve were vigorously investigated and excluded as possible causes of these symptoms. The greatest risk for all cancer patients is the inadequate treatment of their disease, because this is inevitably lethal. The aggressiveness of the therapy and the acceptable risk of complications must therefore be balanced against the risk of recurrence. The neuropathy seems to be closely linked to the development of fibrosis around the nerve trunks. The use of large daily fractions, combined with hot spots from overlapping fields contributed to the severity of the complications.