Background <p>Whether real-time ultrasound guidance changes the minimum local anesthetic dose (MLAD/ED₅₀) of intrathecal ropivacaine with dexmedetomidine for cesarean delivery under combined spinal–epidural anesthesia (CSEA) remains unclear. We compared MLAD between ultrasound-guided and landmark-guided CSEA.</p> Methods <p>In this randomized trial, 60 parturients undergoing elective cesarean delivery received CSEA at the L3–4 interspace with intrathecal ropivacaine plus dexmedetomidine 5&#xa0;µg (ultrasound-guided, <i>n</i> = 30; landmark-guided, <i>n</i> = 30). MLAD was estimated using the Dixon up-and-down method (starting dose 10&#xa0;mg; step size 0.5&#xa0;mg). Hemodynamics, block characteristics, adverse events, intraoperative VAS scores, and neonatal Apgar scores were assessed as secondary outcomes.</p> Results <p>Baseline characteristics were similar. MLAD was higher with ultrasound guidance than with landmark guidance (10.14&#xa0;mg [95% CI 9.87–10.41] vs 9.09&#xa0;mg [95% CI 8.74–9.45]; <i>p</i> = 0.002), confirmed by probit regression (<i>p</i> = 0.004). Compared to baseline values at T0, both systolic and diastolic blood pressure decreased in both groups; however,&#xa0;the reduction at 5 min after injection was smaller in the ultrasound-guided group than in the landmark-guided group. Sensory/motor block profiles, adverse events, intraoperative VAS scores, and Apgar scores were comparable between groups.</p> Conclusion <p>When performing ultrasound-guided CSEA at L3–4 for cesarean delivery, clinicians should anticipate a modestly higher intrathecal ropivacaine requirement than with the landmark technique. Both approaches provide comparable anesthesia quality and maternal–neonatal safety; therefore, intrathecal ropivacaine dosing recommendations should be technique-specific.</p> Clinical trail registration number <p>ChiCTR2200058312</p> Graphical Abstract <p></p>