A total of 241 patients with early breast cancer had serial bone scans and chest radiographs during the first two years of follow-up after primary treatment. Each patient had had a modified radical mastectomy and been found to have involved axillary nodes. They were part of a prospective randomized trial testing the use of L-phenylalanine mustard L-PAM as adjuvant chemotherapy. During the two years, these patients had a total of 832 serial bone scans and 1091 serial chest radiographs. Twenty-five patients (10·4 per cent) had bone metastases detected on sequential scanning, only 13 of whom, however, were asymptomatic at the time of the positive scan. Twelve (5 per cent) patients were found to have pulmonary metastases on routine sequential chest radiography of whom only 8 were asymptomatic at the time of the positive chest radiograph. It is concluded from this study that, apart from their usefulness in the monitoring of clinical trials, serial bone scans and chest radiographs cannot be recommended routinely in the follow-up of asymptomatic patients because of the low yield and high cost involved.