A 50-yr-old man with a history of 19 yr of work in the aluminum smelting industry, including 14 years in the potrooms, was found to have diffuse interstitial fibrosis, slightly more severe in the upper zones. He died of respiratory insufficiency 5 yr after initial presentation. Analysis of lung by electron optical techniques revealed 15,000,000,000 nonfibrous particles and 1,300,000,000 fibrous particles of aluminum oxide/g dry lung, values representing approximately a 1,000-fold increase over background exposure. The nonfibrous particles had a geometric mean diameter of 0.4 mu, and the fibers had a geometric mean length of 1.0 mu, a width of 0.06 mu, and an aspect ratio of 16. X-ray diffraction demonstrated alpha but not gamma aluminum oxide. These studies indicate that previous suggestions relating aluminum-induced fibrosis to the presence of gamma aluminum oxide are not correct. Although pulmonary fibrosis in this case may be a response to a very high total aluminum particle burden, the presence of large numbers of fibers raises the possibility that fibers play a role in aluminum fibrosis.