BACKGROUND: Amphotericin B deoxycholate (AmB-D) is standard of care treatment for neonatal invasive candidiasis (IC). Micafungin (MCA) has broad-spectrum fungicidal activity against Candida spp. We compared the efficacy and safety of intravenous MCA with intravenous AmB-D and assessed the pharmacokinetics of MCA in infants >2-120 days of age with proven IC in a phase 3, randomized, double-blind, multicenter, parallel-group, noninferiority study (NCT00815516). METHODS: Infants were randomized 2:1 to MCA (10 mg/kg/d) or AmB-D (1 mg/kg/d) for ≥21 days. Primary efficacy endpoint was fungal-free survival (FFS) 1 week after last study drug dose. MCA population pharmacokinetics included simulated area under the curve (AUC) at steady state and maximum plasma concentration after 2-hour infusion. AUC pharmacodynamic target exposure was 170 µg·h/mL. RESULTS: Thirty infants received MCA (n = 20) or AmB-D (n = 10). The trial was terminated early because of slow recruitment. FFS was observed in 12 of 20 [60%; 95% confidence interval (CI): 36%-81%] MCA-group infants and in 7 of 10 (70%; 95% CI: 35%-93%) AmB-D-group infants. The most common treatment-emergent adverse events were anemia [MCA: n = 9 (45%); AmB-D: n = 3 (30%)] and thrombocytopenia [n = 2 (10%) and n = 3 (30%), respectively]. Model-derived mean AUC at steady state for MCA was 399.3 ± 163.9 µg·h/mL (95% prediction interval: 190.3-742.3 µg/mL); steady state and maximum plasma concentration after 2-hour infusion was 31.1 ± 10.5 µg/mL (95% prediction interval: 17.0-49.7 µg/mL). MCA exposures were above the AUC pharmacodynamic target exposure. CONCLUSIONS: Within the study limitations, infants with IC treated with MCA achieved similar FFS compared with AmB-D. Both agents were safe and well tolerated.