During the period July 1983 through December 1984, aminoglycoside-resistant Acinetobacter calcoaceticus var anitratus (ACA) were isolated from 98 patients in a university hospital. Eighty-seven percent of patients (85/98) acquired aminoglycoside-resistant ACA in the intensive care unit (ICU) and 92% (90198) of all initial isolates were from sputum. ICU patients with respiratory colonization/infection with aminoglycoside-resistant ACA were compared with matched ICU controls with other gram-negative rods in sputum. Compared with controls, the duration of ICU stay prior to colonization/infection with aminoglycoside-resistant ACA was significantly longer for cases (14.7 days υ 5.9 days, P = 0.002). Although exposures to devices and procedures were not significantly different for the two groups, cases received respiratory therapy significantly longer than did controls (14.7 days υ 6.6 days, P = 0.006). Prior to isolation of aminoglycoside-resistant ACA in sputum, cases received more cephalosporins than did controls (1.9 υ 1.2, P = 0.018); aminoglycoside usage in the two groups was comparable but cases tended to have received aminoglycoside for longer durations before colonization/infection than had controls (9.0 days υ 6.1 days, P = 0.08). Following sputum isolation of ACA, 6 of 22 cases developed ACA bacteremia compared with bacteremia in 2 of 22 controls. We conclude that factors predisposing to colonization/infection with aminoglycoside-resistant ACA were extended ICU care, prolonged respiratory therapy, and prior therapy with cephalosporins and aminoglycoside. In addition, ACA may be a more common cause of secondary bacteremia than previously appreciated.