The influence of steroids on the antibody response to a MF59-adjuvanted influenza vaccine in elderly COPD patients has not been studied previously. In the influenza season 2001/02 (October-February) elderly COPD patients were recruited at 14 doctor's offices and our 250-bed hospital. Patients were stratified into three groups according to current treatment regimen: (a) > 10 mg of prednisolone/day (SS); (b) inhaled steroids (IS); (c) no steroid treatment (control group, CG). All patients were vaccinated with the MF59-adjuvanted influenza vaccine. Antibodies against the influenza strains A/H1N1, A/H3N2, and B were measured at baseline, 4 and 24 weeks after vaccination by hemagglutination inhibition (HI) assay. One-hundred and sixty-two patients completed the study (CG n = 42; IS n = 87; SS n = 33). Mean age was 71.3 years (range 60-89). Twenty-one percent of all patients reported local reactions; no serious adverse events were observed. Four weeks after vaccination, mean geometric HI titres (GMT) for A/H1N1, A/H3N and B increased significantly in all groups (p < or = 0.05). After 24 weeks, GMTs to A/H1N1 and A/H3N2 returned to baseline, while GMTs to type B remained significantly higher than baseline in all groups. Significant differences between the groups as regards GMTs, seroconversion (56-89%) or seroprotection rates (64-93%) were not observed. Systemic steroids did not influence the antibody response towards the MF59-adjuvanted influenza vaccine. We found that the strains included in the vaccine showed varying long-term immunogenicity.