The study aimed to assess the efficacy and safety of brostallicin, a new DNA minor groove binder, at a dose of 10mg/m2, intravenous (i.v.) every three weeks, in patients with advanced or inoperable soft tissue sarcoma (STS) and gastrointestinal stromal tumour (GIST) failing first line therapy. Two groups were recruited: (1) GIST following treatment with imatinib; (2) other STS following treatment with single agent doxorubicin or ifosfamide or a single line of combination therapy. The primary end-point was overall response rate (ORR) as defined by response evaluation criteria in solid tumours (RECIST). Progression free survival (PFS) was a secondary end-point. In the GIST group, a Simon two step design was planned: first step 18 patients, total 32 patients (p1=20% p0=5% alpha=beta=0.1). In the non-GIST group, planned sample size was 40 in a standard Fleming one-step design (p0=10%, p1=25%, alpha=beta=0.1). Forty-three patients with non-GIST and 21 patients with GIST were recruited. In general, the drug was well tolerated. Common Toxicity Criteria (CTC) grade 3 or grade 4 toxicity was granulocytopenia: 70% of patients, 50% of cycles; fatigue: 25% of patients, 8% of cycles; febrile neutropenia: 14% of patients, 4% of cycles. There was one confirmed toxic death due to neutropenic septicaemia. Three patients had clinically significant allergic reactions in 249 cycles delivered. In the GIST group, no patients had a confirmed response and recruitment was discontinued at the first step. In the non-GIST group, there were two confirmed partial responses. The 3 month PFS was 46% in the non-GIST group and 33% in the GIST group. In the non-GIST group, this PFS is in the range of other agents considered active in STS, and may predict for more substantial first line activity. Further investigation in STS other than GIST appears warranted.