Recirculation reduces the efficiency of venovenous extracorporeal membrane oxygenation (V-V ECMO), but patient-specific determinants of recirculation and the feasibility of recirculation-guided flow adjustment are not well defined. In this prospective, single-center pilot study, 21 patients with severe acute respiratory distress syndrome (ARDS) on high-flow V-V ECMO underwent recirculation assessment by ultrasound dilution measurements. Clinically relevant recirculation was defined as ≥10%. Right ventricular dysfunction (RVD) was adjudicated by echocardiography. In patients with relevant recirculation, ECMO blood flow was reduced stepwise by the estimated recirculating blood volume per unit of time. Recirculation was measurable in 19 patients (90%), with a median of 14% (interquartile range [IQR], 7–25). Twelve patients (57%) had a recirculation ≥10%. Patients with RVD showed a significantly higher recirculation than those without (25% [21–30] vs. 9% [7–15], p = 0.024). Flow reduction was attempted in 12 patients; nine (75%) completed the intervention, while in three (25%), the reduction had to be aborted. Systemic oxygenation was maintained despite flow reduction (arterial oxygen saturation [SaO₂] decreased from 99% [97–99] to 98% [95–98], arterial oxygen tension [PaO₂] from 104 mm Hg [83–130] to 83 mm Hg [71–91]; p < 0.05). Recirculation was frequent during high-flow V-V ECMO and was amplified in patients with RVD. Recirculation-guided flow reduction was feasible in most patients and maintained adequate oxygenation in this trial.