In 1950, Whipple and Harris1 reported 4 cases of Laennec's cirrhosis in which there was bacteremia due to Escherichia coli. Spellberg2 and Beeson3 have commented on infections of the blood stream due to colon bacilli in patients with cirrhosis of the liver. In 1954, members of our group4 noted cases of severe hepatic disease complicated by bacteremia due to E. coli. These were among the 4 such cases separately reported,5 and also were described in our study of 137 patients with bacteremia due to Gram-negative bacilli encountered in a 15-year period extending through 1954.6 Subsequently, other reports of this coincidence of diseases have been made.7-12 Although bacteremia due to Gram-positive microorganisms may complicate severe hepatic disease, our experience to date suggests that it may occur less commonly than Gram-negative bacteremia. Only 1 of 143 patients with bacteremia due to Gram-positive organisms who were