This study was conducted as a prospective, multicenter trial to evaluate the efficacy and safety of micafungin as an empirical therapy for suspected invasive fungal infections (IFIs), including febrile neutropenia (FN), and to evaluate the usefulness of β-D: -glucan (BG) and Aspergillus galactomannan (GM) antigen in patients with hematologic diseases. A total of 121 patients were enrolled and assessed for safety, and 119 were examined for clinical efficacy. The main underlying diseases were acute myeloid leukemia (38.0%), acute lymphoblastic leukemia (18.2%), and malignant lymphoma (18.2%). The median initial daily dose and duration of micafungin treatment were 150 mg/day and 13 days, respectively. The overall response rate for suspected IFIs (n = 119), based on four composite endpoints, including baseline IFI, breakthrough IFIs (proven and probable), survival, and premature discontinuation, was 79.0%. In addition, the response rate for FN (n = 81), based on these four endpoints as well as defervescence during neutropenia, was 39.5%. Breakthrough IFIs (proven, probable, and possible) occurred in five patients during micafungin treatment. All of these patients were positive for either BG or GM before the breakthrough IFIs. The incidence of adverse events (AEs) associated with micafungin was 10.7% and most were mild. The majority of AEs were liver dysfunction. These results indicate the effectiveness and safety of micafungin as an empirical therapy for suspected IFIs, including FN, and the usefulness of monitoring both BG and GM to detect breakthrough IFIs.