OBJECTIVE The use of saliva has many advantages and the concentration of some sex steroids in saliva generally reflect serum levels. However, recently doubts have been cast on this relationship. This study was undertaken to establish any correlation between serum and salivary progesterone in quiescent and dynamic states. METHODS Serum and salivary progesterone were measured in normally cycling (N = 23), pregnant (N = 10) and 5 postmenopausal women on hormone replacement therapy, using non-isotopic enzyme immunoassay on microtiter plates developed in this unit. RESULTS Salivary progesterone was found to be approximately 1% of serum progesterone in normally cycling and pregnant women. There was good positive correlation between serum and salivary progesterone in normally cycling, pregnant and the postmenopausal women. The linear regression equation is y = 0.009x + 0.035: r = 0.987, P < 0.001; y = -0.011x -0.079: r = 0.958, P < 0.001; and y = 0.014x + 1.18: r = 0.714, P < 0.001, respectively. The maximum concentration of progesterone in postmenopausal women was achieved simultaneously in serum and saliva at 2 h post therapy thereby supporting the theory of rapid ultrafiltration of steroids from blood to saliva. CONCLUSIONS Salivary progesterone measurement reflects serum concentration and provide an attractive alternative to serum measurement. It should become increasingly popular in gynecological and endocrinological investigations especially in longitudinal studies.