In July, 2005, a 63-year-old non-smoking white woman was admitted to hospital with a 3-month history of nonproductive cough, slightly febrile body temperature, and right upper quadrant abdominal pain. The patient had no history of previous medical problems. On admission, physical examination was normal. Blood test results showed a cholestatic pattern with increased γ-glutamyl transferase (γGT) of 321 U/L and alkaline phosphatase of 204 IU/L. Total serum cholesterol concentration was 7·6 mmol/L. Despite 3 weeks’ antibiotic therapy before admission, C-reactive protein was high at 180 mg/L. Screening for autoantibodies (antinuclear, anti mitochondrial, and anti cytoplasmic antibodies) was negative. Chest radiography on admission showed no abnormalities; high-resolution CT of the lungs showed consolidation of the lingula and the middle lobe region (fi gure A). Fibreoptic bronchoscopy was done; cytological examination of the lavage-fl uid showed predominantly foam cell macrophages (fi gure B) containing large fatty vesicles, and lipid droplets were detected on Sudan III staining, characteristic for lipoid pneumonia. Histological examination of the subsequent transbronchial lung biopsy sample showed mild interstitial infl ammation, without signs of malignancy or other interstitial lung disease. The patient denied mineral oil ingestion or nasal drop misuse. She was therefore diagnosed with endogenous lipoid pneumonia due to hypercholesterolaemia. To clarify the cause of the cholestasis, magnetic resonance cholangio-pancreaticography was done, which showed a narrowed hepatic bile duct system together with vanishing of the small bile ducts—characteristic for primary sclerosing cholangitis. Liver biopsy confi rmed the diagnosis by showing concentric sclerosis around bile ducts with little infl ammatory infi ltration, periductal fi brosis, multifocal strictures, and dilatations. We started our patient on 15 mg/kg ursodeoxycholic acid and 40 mg atorvastatin daily. Her condition improved substantially, and she became completely asymptomatic after 6 weeks. Total serum cholesterol dropped to 4·6 mmol/L and γGT to 100 U/L; alkaline phosphatase and C-reactive protein concentration returned to normal values. CT scan showed marked improvement with complete resolution of lung infi ltration after 6 weeks of treatment. Lipoid pneumonia may be classifi ed as exogenous or endogenous. Endogenous lipoid pneumonia usually develops when lipids that normally are found in the lung tissue escape from destroyed alveolar cell wall distal to an obstructing, usually malignant, airway lesion or from lung tissue damaged by a suppurative process. Less commonly, endogenous lipids appear in the lung with fat embolism or thromboembolism, Wegener’s granulomatosis, pulmonary alveolar proteinosis, or lipid storage diseases. Hypercholesterolaemia is common in patients with primary sclerosing cholangitis and can lead to an accumulation of cholesterol in diff erent tissues and can be the underlying cause of an ongoing infl ammatory process. The lungs are actively involved in lipid metabolism, and macrophages help to clear lipids from the lungs. The mechanisms involved in lipoid pneumonia are unclear but may involve an infl ammatory response that is linked to increased lipid uptake by alveolar macrophages. The activity of the disease is not necessarily mirrored by the serum total cholesterol concentration, which might explain the discrepancy of the severity of disease in our patient and the relatively moderate elevation of the serum cholesterol concentration. Treatment of endogenous lipiod pneumonia is not standardised. Successful treatment with total lung lavage or with systemic steroids has been reported, proposing an underlying autoimmune phenomenon of the disease (supported by our case). Our patient was successfully treated by lowering her serum cholesterol concentration and thereby terminating the underlying cause of infl ammation, strongly supporting the link between endogenous lipoid pneumonia and pathogenic role of serum lipids.