The purpose of this article is to document clinical and pathologic observations concerning liver disease associated with the PI ElembergM phenotype of alpha 1-antitrypsin. Deposits of alpha 1-antitrypsin that were periodic acid-Schiff positive and stained with an antiserum to alpha 1-antitrypsin were found in the liver of a markedly jaundiced, terminally ill patient with Stage IV primary biliary cirrhosis. A biopsy performed three years earlier failed to reveal alpha 1-antitrypsin deposits. The phenotype PI ElembergM was verified by both acid starch gel electrophoresis and isoelectric focusing in agarose. The deposits of alpha 1-antitrypsin in the liver appear to be a consequence of the patient's disease and age and not due to an association with the PI*Elemberg allele. Accumulation of alpha 1-antitrypsin in the liver of this patient may be due to an accelerated synthesis of this protease inhibitor exceeding the liver's capacity for glycosylation or other steps in its secretion.