SUMMARYThe patterns of fractionation used to treat bone metastases in a single centre in two time periods were determined. Clinical audit was carried out, for two periods of 6 months each, in 1988 and 1993. Data recorded included patient, tumour and treatment variables. Palliation of bone metastases represented 40% and 44% of palliative treatment courses, and 20% and 21% of total treatment courses, respectively, in both time periods. Shorter treatment schedules were used in 1993 compared to 1988, with mean fraction numbers of 6.4 versus 8.5, respectively (P<0.001). This reduction in fraction numbers occurred for the common primary tumour sites. In 1993, the mean numbers of fractions used were related to the primary tumour site (7.8 fractions for breast, 5.0 for lung, 6.7 for melanoma, 6.4 for other primary sites,P= 0.05), the treatment site (7.3 fractions for weight‐bearing bones, 4.0 for non‐weight‐bearing bones,P<0.001) and patient address (6.0 fractions for city postcodevs7.8 for country postcode,P= 0.02). Treating radiation oncologist (5.7‐7.8 mean fractions,P= 0.06) and patient age (P= 0.56) were not significant factors for the number of fractions used. It is likely that there were multiple causes for a reduction in the number of fractions used to treat bone metastases, including the results of clinical trials and increasing pressure to optimize the use of scarce resources. However, patients who may have a better prognosis (breast primary) and those with metastases in weight‐bearing bones continued to receive longer treatment