Hypotension after spinal anesthesia for cesarean delivery remains a major clinical problem. Fluid preloading regimens together with vasopressors have been used to reduce its incidence. Previous studies have used noninvasive arterial blood pressure measurement and vasopressor requirements to evaluate the effect of preload. We used a suprasternal Doppler flow technique to measure maternal cardiac output (CO) and corrected flow time (FTc, a measure of intravascular volume) before and after spinal anesthesia after 3 fluid preload regimens. We hypothesized that colloid solutions, compared with crystalloid, would produce the largest increase in CO and have the lowest incidence of hypotension. Sixty healthy term women scheduled for planned cesarean delivery under spinal anesthesia were recruited for this randomized, double-blind study. Baseline heart rate, systolic blood pressure (SBP), CO, and FTc were recorded in the left lateral tilt position. Patients were randomized to receive 1 of 3 fluid preload regimens given over 15 min: 1.5 L crystalloid (Hartman's solution), 0.5 L of 6% w/v hydroxyethyl starch (HES) solution (HES 0.5), or 1 L of 6% w/v HES solution (HES 1.0). Further measurements were made after fluid loading every 5 min for 30 min. After 30 min, spinal anesthesia was induced with hyperbaric bupivacaine 12.5 mg with fentanyl 15 microg and recordings were continued every 5 min for 20 min or until surgery started. The primary outcome, CO, was compared among groups. The incidence of hypotension (defined as a 20% reduction in SBP from the baseline), ephedrine use, and umbilical cord blood gases were also compared. Patient characteristics, heart rate, SBP, and cord gases were similar among groups. Although CO and FTc increased after preload in all groups (P < 0.005), this was only maintained with HES 1.0 after spinal anesthesia (P < 0.005). There were no differences among groups in the incidence of hypotension (70% vs 35% vs 65% for Hartman's solution, HES 0.5, and HES 1.0, respectively; P = 0.069) or mean ephedrine dose (10.4 vs 5.7 vs 9.7 mg; P = 0.26). Despite CO and FTc increases after fluid preload, particularly with HES 1.0 L, hypotension still occurred. The data suggest that CO increases after these preload regimens cannot compensate for reductions in arterial blood pressure after spinal anesthesia.