To the Editor.— Kiang and Kennedy (238:59-60, 1977) have presented an interesting idea that may prove to be useful in the differential diagnosis of adenocarcinomas of unknown primary site. Unfortunately, however, the evidence offered hardly appears to justify their claim that positive estrogen receptor assays can indicate that metastases arose from breast cancer. Of the 295 patients in whom estrogen receptor assays were performed, only 45 patients had cancers of other origin besides breast. A total of six patients with colon cancer is certainly not a substantial sample from which to draw any conclusions, particularly in view of reports such as that of McClendon et al,1who found estrogen receptor protein in five of 21 colonic neoplasms. A review of the five case reports in Kiang and Kennedy's article shows extenuating circumstances that prevent justification of their claim. In case 1, the only evidence that the patient might have