IN GENERAL oestrogen receptor (RE) and progesterone receptor (RP) tend to occur together in breast carcinomas, RP being a marker of RE integrity (Horwitz et al., 1975a). It has also been claimed that RP is associated with a lower risk of distant metastatic spread (Pichon et al., 1980 (Thorsen, 1981) that heparin may interfere with the estimation of RP in vitro. It has further been shown (Hartveit, 1981) Material from the primary tumours had been taken at the time of operation, and divided, one part being used for receptor determination, one for histology. Both receptors were analysed by dextrancharcoal assay (DCA) (McGuire et al., 1975;Horwitz et al., 1975b). Although cytosols were no longer available, a total of 186 paired samples (i.e. RE and RP values and the corresponding histological material) were.From this material we selected all the RP-RE-cases (17), all the RP-RE+ (18) and the first 29 of the RP+ RE+. Values > 3 fmol/mg cytosol were considered positive for RE, > 5 fmol for RP. The age of the patients ranged from 33 to 91 years with a mean of 64. Only 9 of them were under 55. There was no difference in age distribution between the groups.Sections from the formalin-fixed paraffin-embedded blocks from the primary tumours were stained with toluidine blue at pH4 (Hartveit, 1981). The mast cells were counted at the edge of the tumour (in the zone of host-tumour interaction) using light microscopy (x 400). The microscope was connected to a digitizer (Summagraphic: Bit pad one) which in turn was connected to a computer, as described previously (Thoresen et al., 1982). One slide was used per case and the total number of mast cells in 20 fields was counted.The mean values ( s.d.) for the mastcell count were related to the receptor profile of the tumours, as judged from cytosol assay. The profile RP+ RE-is not included, as we had only 3 such cases.The figure shows the range and distribution of the mast-cell counts in the