Background: An altered immune response and decreased vaccine response are observed in patients with chronic renal failure. A preliminary study of 15 non-immunised patients, despite appropriate previous hepatitis B vaccination, showed a 60% seroconversion rate after 3 months of dialysis with a polymethylmethacrylate (PMMA) membrane. This response was associated with circulating soluble CD40 (CD40s) decrease, a natural inhibitor of the humoral immune response. The aim of the study is to confirm these results in a randomised study. Methods: We conducted a multicentre randomised intention-to-treat superiority clinical trial comparing polysulfone and a PMMA membrane in 2 parallel patient groups. The primary end point was the vaccine response rate, as defined by an anti-HBs antibodies titre of >10 IU/L, 1 month after the last vaccination with a double dose of Engerix B20®, performed at weeks 12, 16, 20, and 36. Results: Twenty-five patients were randomised and included in an intention-to-treat analysis. They were dialysed on polysulfone (n = 11) or PMMA (n = 14) for 40 weeks. Fifty percent of the PMMA patients versus 54.5% of the polysulfone patients achieved seroconversion (p = 1.00). The median anti-HBs antibody titre in responders at week 40 was 496 (92–750) versus 395 (43–572) UI/mL for PMMA and polysulfone, respectively (p = 0.46). The median CD40s titre at week 12 was 306 (193–448) versus 491 (281–515) pg/mL (p = 0.21). The CD40s median variation between week 0 and week 12 was 5 (–105 to 90) versus 64 (–63 to 123) pg/mL (p = 0.55). The CD40s level at week 12 in non-responders was slightly inferior to that of the responders: median 193 (168–331) versus 413 (281–512) pg/mL (p = 0.08). Conclusion: We did not observe a better vaccine response with the PMMA membrane compared to high-flux polysulfone. The PMMA membrane did not decrease the CD40s more than the polysulfone membrane probably because the titre was previously low in the 2 groups.