Osmoregulation was studied in near term and age-matched Sprague-Dawley rats. Basal plasma osmolality (P(osm)) and plasma sodium (P(Na)) were 281+/-3 mosmol/kg and 134+/-3 meq/liter, respectively, on the 20th gestational day compared with 292+/-3 mosmol/kg and 140+/-1 meq/liter in virgin animals (P < 0.001), whereas P(urea) and plasma water content were similar in pregnant and control rats. These differences could not be reproduced in animals receiving progesterone, estrone, or a combination of progesterone and estradiol for 2 wk. Pregnant and control rats were deprived of water for periods ranging from 0 to 48 h. P(osm), always lower in gravidity, was 290+/-3 mosmol/kg after 2 d of water deprivation in pregnant animals compared with 300+/-2 mosmol/kg in controls (P < 0.001). Thus 48 h of dehydration were required before P(osm) in gravid rats was similar to basal values in the age-matched virgins. Despite strikingly lower P(osm), plasma arginine vasopressin (P(AVP)) and urinary osmolality (U(osm)) were similar in the basal state averaging 2.16+/-0.78 pg/ml and 1,652+/-406 mosmol/kg, respectively, during pregnancy compared with 2.08+/-2.17 pg/ml and 1,483+/-203 mosmol/kg in controls (NS). Water deprivation increased P(AVP) and U(osm) similarly in pregnant and virgin rats: these values reached 22.7+/-3.3 pg/ml and 3,300+/-123 mosmol/kg at 48 h in gravid compared with 26.0+/-6.4 pg/ml and 3,342+/-141 mosmol/kg in the controls (NS). Regression equations for P(AVP)vs. P(osm) which were highly significant (P < 0.001) in both groups demonstrated an apparent threshold for AVP secretion approximately 11 mosmol lower in gravid animals. Intravascular volume decreased, and plasma aldosterone increased during water deprivation, and both changes (Delta%) were significantly greater in the gravid animals (P <0.01). Therefore, P(osm) was increased without concomitant volume depletion by intraperitoneal hypertonic saline. Again P(AVP)vs. P(osm) correlated significantly (r > 0.9; P < 0.001) in each group, and the apparent threshold was 14 mosmol lower in pregnant animals. Diluting ability, tested by oral water loading, was not impaired in the pregnant animals which excreted a 30 ml/kg load as well as controls. Also, chronically hydrated virgin animals whose fluid intake was more than twice that of pregnant rats (for 19 d) did not lower their P(osm). In separate studies homozygous Brattleboro rats, which produce no endogenous vasopressin, were also shown to have a decreased P(osm) (pregnant 292+/-4 mosmol/kg; virgin 310+/-6 mosmol/kg P < 0.001), but unchanged U(osm) during pregnancy. Data demonstrate a resetting of the osmostat in gravid Sprague-Dawley rats as P(osm) and the threshold for AVP secretion both decrease significantly during gestation in this species. Studies in homozygous Brattleboro animals with hereditary diabetes insipidus suggest that the osmotic threshold for thirst is reset as well.