Background: Bias in industry-initiated trials is common and some journals try to reduce it by requiring manuscript authors to declare that they had full access to the data. But the word ‘‘data’’ is open to interpretation and access does not equate usage. Objectives: To investigate which type of data academic corresponding authors had access to in industry-initiated trials and how they used this access. Methods: We included a sample of industry-initiated randomised trials published in The Lancet in 2008 or 2009 and corresponding trial protocols provided by The Lancet. For each protocol and published paper, we extracted information on data management, analysis, ownership of data and access to data. We will also ask the corresponding authors which type of data they had access to and whether and how the access was used. Wewill compare the information from protocols, published papers and the author survey. Results: We identified 169 papers of randomised trials published in The Lancet in 2008 or 2009 and included 68 industry-initiated trials. An additional 13 industry-sponsored trials that apparently had independent data management and analysis will be analysed separately. We retrieved trial protocols for all 81 trials. Results will be presented at the Colloquium. Conclusions: Awaits results from data analysis.