Experiments were done to examine whether rat relaxin is dipsogenic and whether such dipsogenic effects of rat relaxin are related to time of injection during the light-dark cycle. Female rats were fitted with a chronic intra-cerebro-ventricular (i.c.v.) cannula. Rat relaxin (2.5, 5, 10, 25, 50, or 100 ng/2 microl in 0.9% saline) was injected into the right lateral ventricle at either morning (0800-1000 h), afternoon (1400-1600 h), or night (2200-2400 h), and water consumption was measured. Relaxin caused a dose-dependent dipsogenesis at doses > or = 5 ng, but the sensitivity and magnitude of the response varied with the photoperiod. Water consumption was smallest (3.5 +/- 0.7 ml at 50 ng) and least sensitive (minimal effective dose at 25 ng) in the afternoon and maximal (17.7 +/- 2.3 ml at 50 ng) and most sensitive (minimal effective dose 5 ng) at night. The latency from injection to drinking was 55.8 +/- 10.4 sec (mean +/- SEM) and did not vary significantly with either the dose or time of day. A second set of experiments was done to examine the effects of neutralizing the central actions of relaxin on drinking behavior in pregnancy. Pregnant rats were injected daily, through a chronically implanted i.c.v. cannula, with either a specific monoclonal antibody raised against rat relaxin from day 12 to day 22 of gestation or with saline as a control. Drinking and eating behavior and weight gain were monitored every 12 h during pregnancy. There was a significant decrease in water consumed at night, but no effect on drinking during the day in relaxin-neutralized rats. These animals also showed a decrease in weight gain during pregnancy compared with controls and gave birth to lighter-weight litters. These data provide evidence that the dipsogenic response to exogenous rat relaxin in female rats varies with time of injection during the light-dark cycle and suggest that relaxin in the brain may have a role in nighttime drinking behavior during the second half of pregnancy.