OBJECTIVES: Caspofungin was evaluated as first-line monotherapy of invasive aspergillosis (IA) in patients with haematological malignancies and undergoing autologous transplants. METHODS: Adults with proven or probable IA, defined strictly according to EORTC-MSG criteria, were eligible. Those with possible IA were enrolled, but were not evaluable for efficacy unless upgraded to proven/probable disease within 7 days of registration based on investigations performed within 48 h after enrolment. Caspofungin dosage was 70 mg (day 1) followed by 50 mg/day. The primary endpoint was the proportion of patients with complete or partial response at the end of caspofungin therapy in the modified intention to treat (MITT) group; secondary endpoints were response and survival at day 84 and safety. RESULTS: In the MITT group (n = 61), 75% of patients had cancer not in remission (relapsing or refractory), 85% were neutropenic at enrolment and 49% had a Karnofsky score of < or =50. At end of treatment, 1 and 19 patients had complete and partial response, respectively [success rate 33% (20/61)], 9 (15%) achieved stabilization and 31 (51%) had disease progression. One patient was not evaluable. The 6 and 12 week survival rates were 66% (40/61) and 53% (32/60), respectively. Baseline characteristics associated with survival at day 84 were an underlying disease in remission (not relapsing or refractory) and Karnofsky score. Recovery from neutropenia at the end of treatment was also significantly associated with survival. No serious drug-related adverse events or discontinuations due to drug-related adverse events were observed. CONCLUSIONS: Caspofungin provided an observed response rate compatible with the null hypothesis of a true response rate of < or =35%. Underlying disease-related factors had a major impact on results.