A 11.20 mg dose of progesterone was administered by nasal spray (NS) to six healthy postmenopausal women. Serial blood samples were collected and plasma progesterone was assayed by radioimmunoassay (RIA) according to three different procedures. In the first, RIA was carried out directly on plasma aliquot (Method A), in the second after diethyl ether extraction (Method B) and the third, after diethyl ether extraction and Celite column chromatography (Method C). The mean serum peak level (CMax) calculated with Method A (2.87 +/- 1.14 ng/ml) was higher than that obtained with both Method B (2.24 +/- 0.76 ng/ml) and C (1.58 +/- 0.76 ng/ml; P < 0.05); similarly the area under the curve (AUC) measured with Method A (695.79 +/- 348.24 ng h/ml) was higher than that obtained with both Method B (390.12 +/- 95.16 ng h/ml) and C (243.71 +/- 82.97 ng h/ml; P < 0.02). On the other hand, progesterone serum levels measured with Method C peaked earlier than those observed with Methods B and A (21.67 +/- 19.40, 25.83 +/- 18.55 and 35 +/- 20.70 min, respectively). These data are consistent with the high specificity of Method C for progesterone whereas the other methods could overestimate the progesterone serum levels probably measuring also progesterone metabolites particularly 5 alpha- and 5 beta-dihydroprogesterone. This study confirmed the rapid absorption of progesterone across the nasal mucosa avoiding the first-pass liver metabolism; however, a 'first-pass effect' of the nasal mucosa should be taken into consideration when progesterone is delivered by the nasal route because probably a significant portion of progestational effects are due to its active metabolites.