PRL has a definite activity in the induction and promotion of mouse and in the growth of rat mammary tumors. We and others have found that human PRL or growth hormone (GH) had a growth promoting effect on human mammary cancer cells. It has been shown that prolactin receptors (PRL-R) which are specific for all lactogenic hormones (hPRL, hGH, hPL) are present on mammary cancer cells in long-term tissue culture and also in tumor biopsies. We found that 43% of the tumors had free PRL-R (FPRL-R) and that 72% had total PRL-R (TPRL-R) which have been desaturated in vitro. A significant correlation (Spearman test) was found between PRL-R (especially TPRL-R) on the one hand, estradiol (P less than 0.001) and progesterone receptors (P less than 0.01) on the other. The demonstration of PRL-induced proteins (PIP) might be a better sign of PRL sensitivity than the existence or PRL-R; PIP have been found by Northern blot analysis in 47% of 70 breast cancers. Overall survival (OS) and relapse-free survival (RFS) analysis with a median duration of follow-up of 5.3 yr showed that TPRL-R had a significant prognostic value only in node positive patients (chi 2 = 5.61, P = 0.02). Neither FPRL-R or TPRL-R were a significant prognostic factor when studied by Cox analysis. This confirms our previous results. Since at least some human mammary cancers appear to be PRL-dependent we carried out a multicenter randomized trial comparing as the first hormonal treatment tamoxifen (TAM) (30 mg/day) + bromocriptine (B) (5 mg/day) vs TAM + placebo. 171 patients entered this trial. No difference was observed between the two groups in response rates, duration of response or survival. Recent studies are thus in favor of a role of lactogenic hormones during the course of breast cancer. However no improvement in therapy has been observed yet. The combination of drugs to achieve a total anti-lactogenic treatment, the use of anti-PRL-R antibodies are interesting areas of research; the recent cloning of PRL-R and GH receptors may open new clinical perspectives.