There are limited data on the relationship between voriconazole levels and clinical outcomes relative to invasive aspergillosis (IA). We therefore analyzed the association between initial voriconazole trough blood levels and clinical responses of IA in patients with hematologic malignancies. All adult patients treated with voriconazole in a tertiary care hospital in Seoul, South Korea, between August 2009 and April 2011 were identified from pharmacy records. Initial voriconazole trough levels were routinely measured 1 week after therapy and patient responses were classified as success (complete or partial response) or failure (stable response, progression of disease, or death) at 2 weeks post-therapy. Fifty-two patients, involving 2 (4%) proven, 29 (56%) probable, and 21 (40%) possible IA infections, were included. Of these, 11 (21%) had initial voriconazole levels of ≤ 2 mg/l and the remaining 41 (79%) had > 2 mg/l. There were slightly fewer successful responses (45%, 5/11) in the patients with initial voriconazole levels ≤ 2 mg/l than in those with voriconazole levels > 2 mg/l (51%, 21/41), but the difference was not statistically significant (P = 0.73). Neutropenia (OR 0.1, P = 0.008) and immunosuppression (OR 0.1, P = 0.004) were independently associated with 2-week successful response after voriconazole therapy. In conclusion, initial voriconazole trough levels may not significantly affect clinical outcomes of IA at 2 weeks after voriconazole therapy in patients with hematologic malignancies. Further studies of prospective design are needed to establish the optimal procedure for voriconazole drug monitoring.