The aim of our retrospective study was to compare the application of regional citrate anticoagulation and citrate-related side-effects in plasma exchange (PE) with different replacement solutions. We included 35 patients treated with PE with regional citrate anticoagulation and divided them into three groups according to the replacement solution used: human albumin (HA) group (40 PE treatments), fresh frozen plasma (FFP) group (86 PE treatments), or a combination of the two (63 PE treatments). The citrate anticoagulation parameters, ionized calcium and metabolic consequences of citrate were compared. The blood flow and citrate infusion rates were similar in all groups. To maintain comparable values of ionized calcium during PE, significantly more calcium was replaced in the combination group (7.6 +/- 1.3 vs. 6.2 +/- 2.7 mL/h, P < 0.001) and even more in the FFP group (10.8 +/- 1.7 vs. 6.2 +/- 2.7 mL/h, P < 0.001) as compared to the HA group. The pH increased significantly and comparably in all groups, but the increase in bicarbonate was significantly higher in the FFP group (4.4 +/- 3.0 vs. 2.6 +/- 2.1 mmol/L, P = 0.01). A short, heparin-free hemodialysis session was performed after the PE treatment, because of significant metabolic alkalosis (mainly with pH > or = 7.5), significantly more often in the FFP group (14/86 PE, P < 0.01) as compared to the HA group (0/40), and only rarely in the combination group (2/63). To conclude, when FFP is used as a replacement solution during PE with citrate anticoagulation, significantly more calcium needs to be replaced and the increase in bicarbonate is greater during PE. The additional citrate contained in FFP, combined with frequent PE treatments, often causes significant metabolic alkalosis, which can be efficiently corrected with a short heparin-free hemodialysis.