Abstract Background and Aims The combined use of a heparin-grafted membrane with a citrate enriched dialysate is an effective hemodialysis strategy with low circuit clotting rates while avoiding systemic anticoagulation. Whether this technique results in dialysis efficacy that is non-inferior to regular hemodialysis using systemic anticoagulation has not been investigated up to now. Method Prevalent hemodialysis (n=26) patients were recruited for a randomized crossover non-inferiority trial powered at >90% to detect a prespecified non-inferiority threshold of 10% spKt/Vurea (NCT03887468). Hemodialysis using a heparin-grafted dialyzer in combination with a 1.0 mmol/L citrate enriched dialysate (“evocit”) was compared to hemodialysis using a heparin-grafted dialyzer, systemic unfractionated heparin and regular bicarbonate-based dialysate (“evohep”). Each treatment arm lasted 4 weeks with a 3x4hours weekly hemodialysis regimen. All sessions were standardized with fixed blood- and dialysate flow rates. Biological analyses were done during midweek sessions. The primary endpoint was spKt/Vurea. Secondary endpoints included alternative adequacy markers, premature treatment termination, retransfusion failure and loss of total cell volume of the dialyzer after dialysis. Results A total of 617 hemodialysis sessions were performed: 307 sessions according to evocit and 310 sessions according to evohep protocol. Mean spKt/Vurea was 1.45±0.25 for evocit sessions and 1.50±0.26 for evohep sessions. In a paired analysis, mean of the difference in spKt/Vurea between both study arms was 0.05 with a 95%CI of 0.012-0.098 (p=0.01), the upper bound of the estimate lying within the prespecified non-inferiority threshold (i.e. <0.15). Processed blood volume was 75.4±3L vs 75.8±1.5L and online Kt was 47.3±5L vs 48.3±4L for all evocit and evohep sessions respectively. Urea reduction rate (RR) was 71.3±5.7 vs 72.3±5.8, bèta2microglobulin RR 37.1±8 vs 37.9±8 and myoglobin RR 30.9±9.8 vs 34.5±12.5 for midweek evocit and evohep sessions respectively. Circuit thrombosis leading to premature treatment end occurred in 13/307 (4.23%) of evocit sessions in 6/26 patients but in none of the evohep sessions (p=0.03). Treatment time of evocit sessions complicated with circuit thrombosis (n=13) was reduced with 36 minutes (IQR 20-46 minutes) without impact on effective treatment times overall (236±12 vs 238±4 minutes for evocit and evohep sessions respectively). Retransfusion failure occurred in 3/307 (0.98%) of evocit sessions and none of the evohep sessions. Dialyzers’ total cell volume was reduced with 17% (IQR 11-33%) and 9% (IQR 6-17%) (p<0.0001) after evocit and evohep sessions respectively. Conclusion Hemodialysis avoiding systemic anticoagulation using a heparin-grafted dialyzer with a citrate enriched dialysate is an adequate technique for maintenance hemodialysis offering spKt/Vurea results within recommended dose, and is not inferior to standard hemodialysis using systemic anticoagulation with heparin in terms of spKt/Vurea. Circuit clotting complications occurred at low frequency during evocit sessions and did not have clinically significant repercussions on dialysis efficacy.