Background: A defined bacterial mixture could be a safer alternative to faecal microbiota transplantation (FMT). Aims: To compare the efficacy of a 12-strain mixture termed rectal bacteriotherapy with either FMT or vancomycin for recurrent Clostridioides difficile infection (CDI) in an open-label 3-arm randomised controlled trial. Methods: We screened all individuals positive for C difficile from May 2017 to March 2019. Persons with laboratory-confirmed recurrent CDI were included. Before FMT and rectal bacteriotherapy, we pre-treated with vancomycin for 7-14 days. Rectal bacteriotherapy was applied by enema on three consecutive days and FMT by enema once with possible repetition for two to three infusions within 14 days. The vancomycin group was treated for 14 days with additional five weeks of tapering for multiple recurrences. The primary outcome was clinical cure within 90 days. A secondary outcome was 180-day all-cause mortality. Results: Participants in the FMT group (n = 34) were cured more often than participants receiving vancomycin (n = 31), 76% vs 45% (OR 3.9 (1.4-11.4), P < 0.01) or rectal bacteriotherapy (n = 31), 76% vs 52% (OR 3.0 (1.1-8.8), P = 0.04). Rectal bacteriotherapy and vancomycin performed similarly (P = 0.61). The mortality rate was 6% in the FMT group, 13% in the bacteriotherapy group and 23% in the vancomycin group. FMT tended to reduce mortality compared with vancomycin, OR 0.2 (0.04-1.12), P = 0.07. Conclusions: Rectal bacteriotherapy appears as effective as vancomycin but less effective than 1-3 FMTs. FMT by enema with 1-3 infusions is superior to vancomycin for treating recurrent C difficile infections and might reduce mortality.