The gestational increase in glomerular filtration rate that occurs in the normal rat is exclusively the result of an increase in renal plasma flow and there is no sustained increase in glomerular capillary blood pressure during a normal pregnancy. The factor or factors that initiate the gestational renal vasodilatation (and plasma volume expansion) are maternal, not fetoplacental in origin. The precise nature of the initiating factors has not yet been defined, although it is unlikely that the gestational plasma volume expansion can be the sole cause of the increased glomerular filtration rate seen in pregnancy. A number of vasoactive hormones are activated in pregnancy but as yet no clear candidate has emerged as 'the renal vasodilator'. Preliminary evidence suggests that nitric oxide may play an important role in gestational vasodilatation. The normal kidney in pregnancy exhibits substantial renal reserve to amino acid infusion and unimpaired autoregulatory ability despite being already vasodilated by the gestational stimulus. There are marked and sometimes contradictory changes in the various volume sensing and control systems in pregnancy. In general, the sensors perceiving and controlling intravascular volume are reset during a normal pregnancy to enable to mother to accommodate the increased plasma volume without provoking a natriuretic response. Whether the expanded plasma volume of pregnancy is perceived as normal or underfilled is not clear at this time and may vary according to the volume regulatory system. Repetitive pregnancies do not have any cumulative, long-term deleterious effects on renal function, when the underlying function is normal, when it has been compromised by removal of renal mass or during chronic systemic hypertension in the spontaneously hypertensive rat. In the short term, pregnancy does not worsen kidney function when underlying glomerular damage is due to immune stimuli, ablation of renal mass or gentamicin, or in the spontaneously hypertensive rat. Therefore, the chronic renal vasodilatation of pregnancy does not appear to be a damaging entity, unlike other states of low preglomerular arteriolar resistance, studied in the male rat. When pregnancy is superimposed on Adriamycin nephrosis or chronic blockade of nitric oxide, hypertension occurs and renal function declines. In both situations endothelial damage/dysfunction occurs, as is also seen in pre-eclampsia. Further study of the effects of pregnancy in animal models of endothelial dysfunction will prove rewarding.