Healthy post-menopausal women were randomly assigned to treatment groups receiving 28 day treatment cycles of oestradiol valerate (2 mg, days 1-28) sequentially combined with levonorgestrel (75 micrograms, days 17-28) (n = 19); oestradiol valerate (2 mg, days 1-28) continuously combined with cyproterone acetate (1 mg, days 1-28) (n = 18), or placebo (n = 22). Treatment continued for 2 years. After therapy, the women receiving the sequential combination had an early secretory phase (74%) or atrophic endometrium (26%). All women receiving the continuous combination or placebo had inactive or atrophic endometrium, or the tissue was inadequate for assessment. Biochemical parameters of endometrial secretion [oestradiol dehydrogenase (EDH), isocitrate dehydrogenase (ICDH)] measured in endometrial tissue and placental protein (PP14) measured in serum were all low in the groups receiving the continuous combination and placebo. During sequential treatment, all biochemical parameters showed significantly higher mean values. Serum PP14 accurately reflected this pattern. Furthermore, in the sequentially treated group, serum PP14 correlated highly significantly with EDH (r = 0.70; P less than 0.001) and ICDH (r = 0.57; P = 0.01). These correlations were of the same magnitude as the correlations between the endometrial markers, which indicates that serum PP14 reflects the endometrial status. Furthermore, serum PP14 measurements at 3 months and at 2 years of treatment were highly correlated, reflecting the long-term validity of the measurement. The present data suggest that PP14 measured in serum may be used to assess endometrial status. Further studies are required to determine whether serum PP14 will discriminate pathological endometrial conditions.