Reported central hemodynamics obtained with a Swan-Ganz pulmonary artery thermodilution catheter in preeclamptic patients show marked disparity, which has been interpreted to indicate a variable hemodynamic expression of the disease. However, the variability also may be due, at least in part, to the pharmacological treatment that most of the women studied received during Swan-Ganz measurements. To evaluate the effects of treatment on hemodynamics, we compared the results of Swan-Ganz measurements in 87 preeclamptic women who had received no treatment at all with those obtained in 47 preeclamptic women who had received various drugs and intravenous fluids. Control values were obtained in 10 normotensive pregnant volunteers. Measurements were performed between 25 and 34 weeks of gestation. The median (range) cardiac index in the untreated patients of 3.3 (2.0-5.3) l.min-1.m-2 was significantly lower than that in the treated patients of 4.3 (2.4-7.6) l.min-1.m-2 and in the normotensive pregnant women of 4.2 (3.5-4.6) l.min-1.m-2. The systemic vascular resistance index in the untreated group of 3,003 (1,771-5,225) dyne.sec.cm-5.m2 was significantly higher than that of 2,212 (1,057-3,688) in the treated and of 1,560 (1,430-2,019) dyne.sec.cm-5.m2 in the normotensive control group. The median (range) pulmonary capillary wedge pressure in the untreated group was 7 (-1-20) mm Hg and did not differ from that of 7 (0-25) mm Hg in the treated group. Variability of all hemodynamic variables was much lower in untreated than in treated patients. These results indicate that untreated preeclamptic patients show a rather uniform pattern of a low cardiac index, a high systemic vascular resistance index, and normal filling pressures. The reported extremes of the hemodynamic profile in preeclampsia are artifacts due to treatment.