PURPOSE The purpose of this study was to determine whether the distribution of sensitivities in breast cancer patients, measured using a DNA damage assay on lymphocytes, is likely to provide sufficient discrimination to enable the reliable identification of patients with abnormal sensitivities. MATERIAL AND METHODS Radiosensitivity (x) was assessed in 226 samples of lymphocytes from unselected women with breast cancer and was quantified as the initial number of DNA double-strand breaks (dsb) induced per Gy and per DNA unit (200 Mbp). RESULTS The existence of an inter-individual variation in the parameter (x) is described through the range (0.40-4.72 dsb/Gy/DNA unit) of values found, which have been fitted to the mathematical model defined by the log-normal distribution (mu = 0.42+/-0.03; sigma = 0.52+/-0.03; R(2)=0.9475). A total of 189 patients received radiotherapy after surgical treatment. Among them, we have detected 15 patients who developed severe skin reactions and we have compared their radiosensitivity values with the rest of patients treated. CONCLUSIONS Our results suggest that DNA initial damage measured on lymphocytes offers an approach to predict the acute response of human normal tissues prior to radiotherapy. Values of x higher than 3.20 dsb/Gy/DNA unit theoretically should correspond to the highly radio-sensitive patients. Using the experimental results, we have calculated the strength of the test by means of the area under the receiver operator characteristic curves (A(Z)) to determine whether the radiosensitivity assay can discriminate between patients according to their radiation response. The value found (A(Z)=0.675+/-0.072) is indicative of a fair-poor discriminating capacity of the test to identify the patients with higher risk of developing a severe acute reaction during the radiotherapy treatment.