BACKGROUND/AIMS: Radiotherapy (XRT) is the gold standard treatment for cancer-induced bone pain (CIBP), but only 50% of patients achieve adequate pain relief within 6weeks. No predictors of analgesic response to XRT are known. The aim of this preliminary study was to explore the effect of XRT on sensory changes in CIBP with a view to predicting response. METHODS: After ethics committee approval, patients with CIBP were assessed prior to and 4-6weeks after palliative XRT. This included completion of the Brief Pain Inventory (BPI) and bedside Quantitative Sensory Testing (QST) measuring evoked sensations to quantified stimuli on the skin over the area of CIBP and a control site. RESULTS: Twenty-three patients were assessed pre and post XRT. Thirteen (57%) had an analgesic response (defined as ⩾30% reduction in total BPI). Those patients who had normalisation of abnormal warm sensation ("warm responders", n=6) were different in that they had higher baseline functional BPI pain scores (median score (IQR) in warm responders=43 (31.75-58) compared to 31 (12-39.5) in the remaining patients, p=0.039), larger reductions in pain scores (median difference of 33.5 in total BPI, p=0.027) and increased likelihood of resolution of sensitivity to pinprick. CONCLUSIONS: This is the first clinical study to demonstrate alterations in sensory responses in CIBP. Alterations in specific sensory characteristics seem to be associated with an increased likelihood of successful analgesia from palliative XRT. This supports the use of QST in further biomarker studies to predict response to therapy and aid clinical decision making.