Abstract Background The participant information sheet (PIS) provided to potential trial participants is a critical part of the process of valid consent. However, there is long-standing concern that these lengthy and complex documents are not fit-for-purpose. This has been supported recently through the application of a performance-based approach to testing and improving readability called user testing. This method is now widely used to improve patient medicine leaflets - determining whether people can find and understand key facts. This study applied for the first time a controlled design to determine whether a PIS developed through user testing had improved readability over the original, using a sheet from a UK trial in acute myeloid leukemia (AML16). Methods In the first phase the performance of the original PIS was tested on people in the target group for the trial. There were three rounds of testing including 50 people in total - with the information revised according to its performance after each of the first 2 rounds. In the second phase, the revised PIS was compared with the original in a parallel groups randomised controlled trial (RCT) A total of 123 participants were recruited and randomly allocated to read one version of the PIS to find and show understanding of 21 key facts. Results The first, developmental phase produced a revised PIS significantly altered in its wording and layout. In the second, trial phase 66% of participants who read the revised PIS were able to show understanding of all aspects of the trial, compared with 15% of those reading the original version (Odds Ratio 11.2; Chi-square = 31.5 p