The purpose of this study was to compare meropenem monotherapy with combination therapy for empirical treatment of neutropenic fever in children with lymphoma and solid tumors. Ninety episodes of neutropenic fever in children (0.7-16.0; mean age 7.7 years) with solid tumors in a single center were randomized to receive either meropenem (50 mg/kg/dose-maximum 1 g, every 8 hours) or piperacillin (200 mg/kg/dose, every 6 hours) plus amikacin (15 mg/kg daily). Failure was defined as treatment modification. Non-Hodgkin's lymphoma (NHL) accounted for 62.2 percent of all episodes, and solid tumors (37.8%) for the rest. Blood cultures were positive in 23 percent of all episodes. Sixty-seven percent of all isolated microorganisms stained Gram-positive. Overall success was 70.0 percent (63/90). The success with meropenem was comparable to that seen with piperacillin plus amikacin: 76.6 versus 64.6 percent (p = 0.25). The failure rate was 33 percent with Gram-positive culture and 78 percent with Gram-negative or mixed cultures. The solid tumor group had significantly less bacteremia (4/34 versus 17/56; p < 0.05) and treatment failure (3/34 versus 24/56; p < 0.001) than the NHL group. No serious drug-related adverse event was noticed. Meropenem monotherapy was as effective as piperacillin plus amikacin combination in the empirical treatment of neutropenic fever in children with lymphoma and solid tumors.