We report the occurrence of a high rate of false-positive test results during the surveillance of hematology patients for galactomannan (GM) antigenemia. Among 218 patients surveyed from June 2002 through June 2003, 42 (19.3%) had > or =1 serum sample positive for GM (optical density index, >1.5). Of these patients, 38 had no additional evidence of invasive aspergillosis, and, therefore, their test results were considered to be false-positives. Case-control analysis showed that treatment with piperacillin-tazobactam was the only risk factor significantly associated with receiving false-positive test results. When tested for GM antigen, 3 of 4 piperacillin-tazobactam batches had positive results. Physicians should be aware of the possible interference of treatment with piperacillin-tazobactam when interpreting the results of the GM assay.