Bone scintigraphy (BS) is commonly performed in the staging and postoperative monitoring of breast cancer. Nevertheless, due to low specificity it often demonstrates hot spots with equivocal interpretation, which may be misleading in the management of these patients. The aim of this study was to assess the value of a serum tumour marker panel in selecting among the patients with equivocal BS those with bone metastases. Between January 1986 and December 1995, 297 breast cancer patients were followed-up after mastectomy with serial determinations of a CEA-TPA-CA15.3 tumour marker panel, BS and liver echography. The tumour marker panel was used to select patients with equivocal BS for examination of suspicious bone areas by further imaging techniques. Up to December 1995, 158 (53%) patients showed an equivocal BS and 47 patients developed bone metastases. In the 158 patients with equivocal BS, prolonged clinical and imaging follow-up over 45 months (mean; range 12-120) was used to ascertain the presence or absence of bone metastases. In these 158 patients the negative predictive value and positive predictive value of the tumour marker panel to predict bone metastases was 97% and 75% respectively. This study shows that in breast cancer patients the CEA-TPA-CA15.3 tumour marker panel has a high value in selecting those patients with bone metastases, or at high risk of developing clinically-evident bone metastases, among the large number of subjects with equivocal BS.