Background and objective Severe preeclampsia may be managed expectantly before 34 weeks gestation with close surveillance. Utilized in fetal growth restriction (FGR), evidence supports umbilical artery (UA) Doppler preventing neonatal morbidity from hypertensive disease and predicting adverse outcomes in preeclampsia. We evaluated the association of abnormal UA Doppler waveforms with early delivery (before 34 weeks gestation) and adverse maternal-fetal outcomes in patients with early severe preeclampsia without FGR. Methodology This is a retrospective cohort study of singleton pregnancies with International Classification of Diseases (ICD) Ninth or Tenth Revision, defined severe preeclampsia diagnosed before 34 weeks gestation without FGR from January 1, 2018, through January 27, 2023, at a large tertiary care center where S/D ratios were calculated from UA Doppler interrogation of a free loop of cord at least once weekly. This study was approved by the IRB (ID00002216) and granted a full Health Insurance Portability and Accountability Act (HIPAA) waiver of consent. Exclusion criteria were major congenital anomalies, congenital infection, aneuploidy, leaving against medical advice >24 hours, and patient instability on admission defined as condition(s) precluding expectant management by the American College of Obstetrics and Gynecology. The primary outcome was delivery before 34 weeks gestation. Secondary outcomes were the mode of delivery and maternal/fetal complications. Patient characteristics and outcomes for normal versus abnormal UA Doppler groups were compared with chi-square, t-tests, and Fisher's exact test. Odds ratios and relative risks were calculated to compare outcomes. Results Of 194 patients with severe preeclampsia, 107 met inclusion criteria. Thirty-four patients had abnormal UA Doppler studies. There were no differences in demographic and clinical data between patients with normal and abnormal UA Doppler studies. Patients with abnormal UA Doppler studies were more likely to deliver before 34 weeks (OR=3.91; 95% CI 1.24-12.33) for worsening severe features (OR=3.85; 95% CI 1.42-10.41), and were less likely to deliver vaginally (OR=0.12; 95% CI 0.03-0.54). Abnormal UA Doppler studies were associated with an increased risk of neonatal complications (OR=6.46; 95% CI 1.42-29.42) and respiratory distress syndrome (RDS) (OR=4.75; 95% CI 1.32-17.16). Abnormal UA Doppler subgroups were divided into patients with elevated S/D >95% Acharya (N=22) and absent end-diastolic flow (EDF) (N=10). The elevated S/D group tended to deliver before 34 weeks gestation for worsening severe features (OR=3.71, 95% CI 1.144-12.050) and had a higher risk of neonatal complications (RR 1.404; 95% CI 1.213-1.624). The absent EDF subgroup was more likely to deliver before 34 weeks (RR=1.52; 95% CI 1.29-1.79) for abnormal fetal testing (OR=6.92; 95% CI 1.71-28.08) and undergo primary cesarean delivery (OR=7.23; 95% CI 1.43-36.61). Conclusion Pregnancies with severe preeclampsia without FGR displayed a high incidence of abnormal UA Doppler waveforms associated with loss of clinical stability and adverse fetal outcomes. The groups with more impedance to umbilical artery flow tended to deliver earlier, and as the Doppler shifted from elevated S/D to absent end-diastolic flow, the mode of delivery shifted to cesarean delivery with increased risk of abnormal fetal testing. These results support the utility of UA Doppler surveillance in severe preeclampsia.