Patients with cystic fibrosis (CF) have reduced peripheral muscle strength. We tested the hypothesis that steroid treatment contributes to muscle weakness in adults with CF. Twenty-three stable CF patients were studied. Measurements included knee extensor (KE), knee flexor (KF), elbow flexor (EF), handgrip (HG), expiratory (Pemax), and inspiratory (Pimax) muscle strengths. Spirometry, body mass index (BMI), and days spent in hospital over the preceding 12 mo (DH) were also measured. Average daily dose of prednisolone over the preceding 12 mo (ADD) was 5.1 mg/day. Pearson's correlation analysis revealed that ADD correlated significantly with skeletal muscle strengths (KF%, r = -0.63, P < 0.01) with the exception of HG%. These findings are independent of age, BMI, pulmonary function, and DH. Multiple-regression analysis revealed that ADD was the most significant predictor of all measures of skeletal muscle function except HG%. It was independently responsible for 54% of the variance in Pimax%, for 46% of the variance in Pemax%, for 45% of the variance in KE%, for 39% of the variance in KF%, and for 41% of the variance in EF%. Concomitant medications (e.g., theophylline) were shown to have no causative effect. Corticosteroids contribute to the skeletal muscle weakness seen in CF patients. The correlation of proximal muscle strength, but not HG strength, with steroid dosage further supports a cause-effect relationship.