Seven laboratories of the EORTC Receptor Study Group reported the distribution of oestrogen (ER) and progesterone receptors (PR) routinely assayed in breast cancer cytosols. A low interlaboratory variability was demonstrated for the median values, and for the frequency of positive tumours as measured by enzyme immunoassay (EIA). Larger variations were found for the frequency of positive tumours, as measured by radioligand binding assay (RLA). They are probably due to differences in the cut-off levels and in the sensitivity of the assay. Analysis of the variability over time clearly demonstrated that the ER-EIA values initially increased compared with RLA. A possible source of variations could be the calibration drift in the ER-EIA kit. In conclusion, quality assessment of steroid receptors should be monitored by comparison of both common standards and distributions routinely obtained in each laboratory. In-house analysis over time is also essential for reagent survey.