The recombinant full-length FVIII product Kogenate has been reformulated using sucrose (rFVIII-FS) instead of human serum albumin as a stabiliser in purification and formulation. The in vivo recovery, haemostatic efficacy, and safety of rFVIII-FS were investigated in 20 previously treated patients with severe or moderate haemophilia A for > or = 24 weeks. In vivo recoveries of 73.5 +/- 16.3%, 78.4 +/- 16.1%, and 82.8 +/- 23.9% after the initial infusion of 50 IU kg(-1) rFVIII-FS and at weeks 12 and 24, respectively, showed no significant changes over time. A total of 1115 infusions (mean dose 24.1 +/- 8.4 IU kg(-1)) were included in the analysis of haemostatic efficacy. One (80.5%) or two (8.2%) infusions achieved adequate haemostasis in 88.7% of all bleeding episodes, and haemostatic efficacy was judged 'excellent' or 'good' in 749 of 764 episodes (98.0%). The haemostatic efficacy was judged as 'excellent' or 'good' in 924 of 1115 (82.9%) infusions. Twenty-one adverse events were observed in 12 patients in the total 1541 infusions included in the safety analysis. Causality with respect to rFVIII-FS could not be ruled out in three events in one HIV-negative patient: elevated CD4(%), decreased CD8(%), and elevated CD4/CD8 ratio. No FVIII inhibitor development was observed in any patient. ELISA assay testing for antibodies to rFVIII, baby hamster kidney cell (BHK) protein, and murine IgG were all negative. These results show that rFVIII-FS is a safe and effective for long-term treatment of patients with haemophilia A.