In order to know the secretory behaviors of human growth hormone (hGH), human prolactin (hPRL), and chorionic somatomammotropin (hCS) in normal and molar pregnancy or puerperium, and after evacuation of hydatidiform mole, the following studies were undertaken.<BR>Forty five normal pregnant women of every period of gestation, 40 women of postpartum (good or poor lactation), 20 patients with molar pregnancy between the 13th and 19th gestational week, 25 patients with good prognosis after the evacuation of hydatidiform mole, 20 patients with destructive mole, and five nonpregnant subjects volunteered for this study.<BR>All studies were started at 0800h after an overnight fast. The subjects reported to the endocrine research facility and were placed at bed rest in a quiet room for about one hour. A#18 Medicut was placed in an antecubital vein and kept patent with a slow drip of normal saline. Two baseline blood samples were drawn 15 min. apart and the results were averaged. Immediately after the second sampling, 30g of arginine was administered intravenously over 30 min. Blood samples were collected at 30, 45, 60, 90 and 120 min. after the initiation of the infusion.<BR>Serum was immediately separated from the blood samples by centrifugation of 3000 rpm for 10 min. and stored for assay at-20°C. HGH levels were measured by a radioimmunoassay kit (Dinabot, Tokyo, Japan). Serum hPRL was determined by a double antibody radioimmunoassay system using hPRL supplied by NIH-NIAMDD, U.S.A. HCS level was measured by the hCS-Kobe double antibody radioimmunoassay system. Immunological crossreaction was not observed between these radioimmunoassay systems within the assay ranges. Serum 17β-estradiol and progesterone were measured by a CIS(CEA-IRE-SORIN, France)'s radioimmunoassay kit for each of the steroid hormones. Statistical analysis was carried out using a Student's t test.<BR>1. The secretions of serum hGH, hPRL and hCS during normal pregnancy were as follows : <BR>The baseline level of hGH remained almost unchanged throughout the course of pregnancy, but that of hPRL and hCS increased with the progress of pregnancy.<BR>The hGH response to arginine infusion decreased (p=n.s. in first trimester, P<0.01 in second trimester, P<0.001 in third trimester), but the hPRL response increased along with the progress of pregnancy (p=n.s. in first trimester, PC0.001 in second and third trimester).<BR>During arginine infusion, the hCS concentration did not change.<BR>2. The secretions of serum hGH and hPRL in puerperium were as follows : <BR>The baseline level of hGH did not change as compared to that of nonpregnant or pregnant women.<BR>The hGH response to arginine infusion was suppressed at the first postpartum week (P<0.01), then recovered to the nonpregnant level at the fifth postpartum week.<BR>Basal and response hPRL maintained a high level from the first to third postpartum week, and decreased at the fifth postpartum week (P<0.001-0.01, respectively).<BR>No significant difference was seen in basal and response hPRL among the subjects with good or poor lactation.<BR>3. The secretions of serum hGH, hPRL, hCS and steroid hormones in molar pregnancy were as follows : <BR>The baseline level of hGH did not change as compared to that of nonpregnant or pregnant women.<BR>The hGH response to arginine infusion in molar pregnancy was greater than that in normal pregnancy (P<0.01).<BR>Basal and response hPRL were higher than that of normal pregnant women in the same gestational week (P<0.01).<BR>On an average, hCS concentration in molar pregnancy was 0.22/μg/ml and significantly low compared to that in normal pregnancy (P<0.001).<BR>The mean value of 1713-estradiol was lower in molar pregnancy than in normal pregnancy (P<0.01), but no significant difference existed regarding the levels of progesterone.<BR>4. The secretions of serum hGH,