Prospective data on the efficacy of secondary prophylaxis in adults with haemophilia A are limited. To analyse bleeding outcomes in the sucrose-formulated recombinant factor VIII [rFVIII-FS (control)] arm of the LIPLONG study, a randomized, double-blind, 52-week trial was conducted in patients with severe haemophilia A receiving prophylaxis with the investigational product BAY 79-4980 or rFVIII-FS. The per-protocol population of previously treated patients with severe haemophilia A without a history of inhibitors (n = 68 males; mean age, 34.4 years) received 25 IU kg−1 rFVIII-FS three times per week for a median of 50.7 weeks. Annualized bleeding rates were assessed and analysed according to predefined target joint status at study start, prestudy treatment type (prophylaxis vs. on demand), age (<30 or ≥30 years), geographical region, bleeding frequency during the previous 6 months and physical activity status during the study using the Student t-test. The annualized median (range) number of bleeds was 2.2 (0.0–23) bleeds per year. The median (range) number of bleeds per year was significantly lower in patient subgroups without vs. with target joints [0.5 (0.0–17.1) vs. 4.2 (0.0–22.8); P = 0.02] and in those with ≤9 vs. >9 bleeds during the previous 6 months [1.1 (0.0–19.2) vs. 5.3 (0.0–22.8); P = 0.01]. Following randomization to prophylaxis with rFVIII-FS, bleeding frequency was effectively reduced. Absence of target joints and prestudy bleeding frequency were predictors of a low bleeding frequency during prophylaxis treatment.