To evaluate the effect of adding amikacin (AMK) to imipenem-cilastatin (IPM/CS), we conducted a randomized controlled trial in patients who experienced neutropenia (< 1,000/mm3) and fever (> 38 degrees C) induced by cancer chemotherapy. There were 70 patients who entered the trial; 34 and 36 patients received IPM/CS plus AMK (arm A) and IPM/CS (arm B), respectively. There was no significant difference in patient characteristics between the two groups. Among 67 evaluable patients, 29 of 32 (91%) and 25 of 35 (71%) responded to the antibiotics therapy in arm A and B, respectively, with EORTC criteria (p < .047). Median days of antibiotics administration and of febrile episode over 38 degrees C were not statistically significantly different between arm A and B. There was no patient with severe side effects, such as seizure, and 17 patients (30%) experienced emesis in both groups. These data suggest IPM/CS plus AMK is therapeutically superior to IPM/CS alone in patients with neutropenic fever induced by cancer chemotherapy.