Non-steroidal antioestrogens, such as tamoxifen, inhibit the growth of human breast cancer cells. The experiments described here compare and contrast the efficacy of tamoxifen and the 'pure' antioestrogen, ICI 164384, on the inhibition of proliferation of MCF-7 cells. Previous studies have shown that ICI 164384 has a greater maximal inhibitory effect than conventional antioestrogens on the growth of MCF-7 cells. Both types of compound block progression of cells through the cell cycle in the early G1 phase. These studies have been extended to measure the population distribution of antioestrogen-treated cells by the use of two-parameter flow cytometry. ICI 164384 proved to be more effective than tamoxifen in decreasing the proportion of actively growing cells in an asynchronous population. In cells grown in the complete absence of exogenous oestrogens, growth was stimulated by oestradiol, insulin, insulin-like growth factor-I (IGF-I) or transforming growth factor-alpha (TGF-alpha). The potent metabolite of tamoxifen, trans 4'-hydroxytamoxifen (4'-OHT), alone also stimulated growth, whereas ICI 164384 did not. Oestradiol and insulin added together demonstrated a clear synergistic enhancement of cell growth. Correspondingly, the stimulatory effect of 4'-OHT on growth was magnified in the presence of insulin, and a combination of ICI 164384 with insulin revealed a much weaker stimulatory action of the 'pure' antagonist. For both compounds the interaction with insulin was complex and characterized by a bell-shaped dose-response curve. However, for 4'-OHT at all concentrations in the range 1 pM-1 microM in the presence of insulin, cell numbers were greater than in cultures exposed to insulin alone. This was not the case for ICI 164384 which suggested that differences in efficacy may be due to interactions between oestrogen and growth factor-mediated mechanisms. Furthermore, ICI 164384 was more effective in inhibiting the action of IGF-I and TGF-alpha alone or in combination, although both antioestrogens produced a partial blockade of growth factor responses in the complete absence of oestradiol. It is concluded that the difference in efficacy between partial agonist and 'pure' antagonist antioestrogens to inhibit growth in vitro is consistent with the difference in the pharmacological profile of these compounds. The absence of stimulatory activity of ICI 164384 is of particular significance in reducing to a minimum the synergistic interaction between oestrogens and insulin.(ABSTRACT TRUNCATED AT 400 WORDS)