The 10-year results are presented of a prospective trial of 411 patients with breast carcinoma treated by mastectomy and post-operative radiotherapy given in either 2 or 3 fractions per week (i.e. a comparison of 6 fractions in 18 days with 12 fractions in 28 days). The early radiation effects on the normal tissues were similar and acceptable. The late skin changes in the chest wall (treated with 70 kV X rays) were progressive and by 10 years were slightly more marked with 6 fractions. Late subcutaneous fibrosis in the axilla (treated with cobalt-60 teletherapy), however, was much less in the 6-fraction group. Twelve fractions resulted in greater restriction of shoulder movement and an increased incidence of lymphoedema of the arm. Doses were selected on the basis of past clinical experience. The dose used to treat the axilla in 6 fractions was 35 Gy, 14.99% less than that predicted by NSD. The dose predicted by alpha:beta, using a value of 2 Gy for late reactions, is 38.14 Gy. Thus simple theory, which omits time, still predicts too high a dose for 6 fractions, although it is closer than NSD. In this trial, the 6-fraction technique showed an advantage over the 12-fraction technique. It was equally effective in controlling local recurrence and had fewer late sequelae. It was also convenient for patients and economic in the use of radiotherapy resources.