Many countries are implementing computer science education, including programming, into their school curricula. In primary school this topic is new, both for students and teachers. Even though there has been research into the difficulties that high school and adult students face when learning to program, not all these findings translate well to the cognitive abilities of younger students and the environment they learn in. Our research focuses on finding out which misconceptions about programming primary school children hold and develop. This is important because research from other fields has shown that knowledge about misconceptions can improve teaching. This project aims to develop a Programming Misconceptions Assessment Tool (ProMAT) for children in grades 5 and 6 that have learned to program using either xLogo or Scratch – two popular educational programming languages currently used in Switzerland. Ultimately, the project should result in an assessment tool that can be used both by researchers and practitioners interested in primary school children’s misconceptions about programming. In addition, it should reveal if and how educational programming environments affect the development of these misconceptions. The ProMAT is currently under development and underwent a first pilot testing phase with 57 children. In this paper, we describe the development of the first version of this tool, discuss insights gained from the pilot study, and outline the next steps of the project.