In a randomized crossover trial, we compared a simple citrate anticoagulation protocol for high-flux hemodialysis with standard anticoagulation by low-molecular-weight heparin (dalteparin). Primary end points were urea reduction rate (URR), Kt/V, and control of electrolyte and acid-base homeostasis. Secondary end points were bleeding time at vascular puncture sites and markers of activation of platelets, coagulation, and fibrinolysis. Solute removal during citrate dialysis was excellent (URR, 0.71 +/- 0.06; Kt/V, 1.55 +/- 0.3) and similar to results of conventional bicarbonate hemodialysis anticoagulation with dalteparin (URR, 0.72 +/- 0.04; Kt/V, 1.56 +/- 0.2). Electrolyte control was effective with both anticoagulation regimens, and total and ionized calcium, sodium, potassium, and phosphate concentrations at the end of dialysis did not differ. Alkalemia was less frequent after citrate than conventional dialysis (pH 7.5 in 25% versus 62% of patients; mean pH at end of dialysis, 7.46 +/- 0.06 versus 7.51 +/- 0.07; P < 0.01). Bleeding time at puncture sites was shorter by 30% after citrate compared with dalteparin anticoagulation (5.43 +/- 2.80 versus 7.86 +/- 2.93 minutes; P < 0.001). Activation of platelets, coagulation, and fibrinolysis was modest for both treatments and occurred mainly within the dialyzer during dalteparin treatment and in the vascular-access region during citrate anticoagulation. Citrate-related adverse events were not observed. We conclude that citrate anticoagulation for high-flux hemodialysis is feasible and safe using a simple infusion protocol.