Forty-nine patients with chronic obstructive lung disease (COPD) were block-randomized in four groups to investigate, if different degrees of steroid-load influenced the effect of pneumococcal-vaccination on antibody level and clinical variables during 6 months of follow-up. The groups included 13 patients without systemic steroids for the previous 3 months, all vaccinated at entrance in the study and treated with steroids for 4 weeks. Nine patients had chronic steroid treatment both before and during the investigation; they were vaccinated at the entrance in the study. Fifteen patients without systemic steroids for the previous 3 months were vaccinated after the end of a 4-week steroid treatment. Twelve patients served as controls, and were not vaccinated. Totally, 60%-78% of vaccinated patients in the three groups had a rise in antibody level, and a later decrease compared to two of the 12 control patients (p<0.01). This difference was also significant (p<0.05) for the patients vaccinated at entrance in the hospital. No differences were observed among the clinical variables: pneumonia, exacerbations, admittance to hospital, increase in the use of steroids or beta-agonists, and the use of antibiotics. We conclude that a rise in antibody level after pneumococcal vaccination can be expected in patients with COPD despite of the use of systemic steroids. The clinical effect of vaccination is debatable.