With the increasing emphasis on earlier detection of breast cancer, detecting methods requiring only a small number of tumor cells are needed. This paper presents our experience of detecting the breast cancer ER with 17-fluoresceinated estrone (17-FE). The results were compared with that obtained from the same tumor by biochemical method. There is no direct quantitative relationship between the ER values (fmol/mg cytosol protein) determined by biochemical method and the proportion of ER containing cells observed by 17-FE method either in cells from needle aspirates prior to tumor removal or in cells from the removed tumor. It was found, however, that detecting the ER by 17-FE, all the tumors with ER levels greater than or equal to 10 fmol had greater than or equal to 10% marker positive cells, but 70% of tumors with negative or borderline (less than 10 fmol) ER levels had less than or equal to 10% ER positive cells. The correlation between the results from the biochemical method and those with 17-FE binding using needle aspirates was better than that between the former and the latter when using cells from fragments of removed tumor. The lack of concordance in 17-FE binding between the needle aspirates and parts of the removed tumor may be related to the diverse cellular composition of a tumor and the way of obtaining the sample. Our experience in this and previous studies show that further study and trial are worthwhile.