Objective: To describe the characteristics of randomized controlled clinical trials (RCT) on cancer drugs conducted in Spain between 1999 and 2003 based on their protocols. Methods: We conducted an observational retrospective cohort study to identify the protocols of RCTs on cancer drugs authorized by the Agencia Española del Medicamento y Productos Sanitarios (AEMPS) (Spanish Agency for Medicines and Medical Devices) during 1999-2003. A descriptive analysis was completed and the association between variables based on the study setting and sponsorship were assessed. Results: We identified a total of 303 protocols, which included 176,835 potentially eligible patients. Three-quarter of the studies were internationally-based, 61.7% were phase III, and 76.2% were sponsored by pharmaceutical companies. The most frequently assessed outcomes were response rate (24.7%), overall survival (20.7%), and progression-free survival (14.5%). Of all protocols, 10.6% intended to include more than 1000 patients (mean: 2442, SD: 2724). Compared with their national counterparts, internationally-based studies were significantly larger (p<0.001) and were more likely to implement centralized randomization (p<0.001), blinding of the intervention (p<0.001), and survival as primary outcome (p<0.001). Additionally, most internationally-based studies were sponsored by pharmaceutical companies (p<0.01). In a high percentage of protocols, the available information was not explicit enough to assess the validity of each trial. Compared to other European countries, the proportion of Spanish cancer drugs protocols registered at www.clinicaltrials.gov (7%) was lower. Conclusion: RCTs on cancer drugs conducted in Spain between 1999 and 2003 were more likely to be promoted by pharmaceutical companies rather than by non-profit national groups. The former were more often part of international studies, which generally had better methodological quality than national ones. There are some worldwide on-going initiatives that aim to increase the transparency and quality of future research. © 2013 Bonfill et al.