This phase I trial determined the recommended dose for phase II trials (RD) of carboplatin 1-h intravenous (i.v.) infusion followed by PM00104 1-h i.v. infusion on Day 1 every 3 weeks (q3wk) in adult patients with advanced solid tumors. A toxicity-guided, dose-escalation design was used. Patients were stratified and divided into heavily (n = 6) or mildly pretreated (n = 14) groups. Transient grade 4 thrombocytopenia (in one heavily and three mildly pretreated patients) was the only dose-limiting toxicity (DLT) observed. Carboplatin AUC3-PM00104 2.0 mg/m(2) was the RD in both groups. At this RD, the carboplatin AUC was equal to ~60 % the target AUC used in other combinations, and the PM00104 dose intensity was 56-67 % of the value achieved at the RD for single-agent PM00104 given as 1-h infusion q3wk. Most treatment-related adverse events were grade 1/2. They mainly consisted of gastrointestinal and general symptoms, such as fatigue, anorexia, mucosal inflammation or nausea. Transient neutropenia (50 % of patients) and thrombocytopenia (33-38 %) were the most common severe hematological abnormalities; their incidence was higher than with single-agent PM00104. No pharmacokinetic drug-drug alterations occurred. Partial response was found in one patient with triple negative breast cancer pretreated with paclitaxel/bevacizumab. Three patients with colorectal cancer, head and neck cancer, and tumor of unknown origin had disease stabilization for ≥3 months. In conclusion, no optimal dose was reached due to overlapping myelosuppression despite stratification according to prior treatment. Therefore, this carboplatin plus PM00104 combination was not selected for further clinical research.