This paper reflects on the challenges and pitfalls of doing collaborative ethnography in a research project (ARITHMUS) that studies the enactment of populations through statistics. Successful collaboration is essential in order to translate the idea of a transnational field of statistical practices – the conceptual starting point through which the researchers of the ARITHMUS team seek to overcome methodological nationalism – into a corresponding methodology and research practice that transcend nationally bounded case studies. Hence, the question arises as to how do we make collaboration work in practice? In the first part of this working paper we explain why we seek to transcend methodological nationalism and why the conceptual starting points of the enactment of a European population and a transnational field of statistical practices require what we call a transversal method. In the paper’s second part we reflect on five interrelated pitfalls and challenges of collaboration in order to tease out possibilities for negotiating these to make collaboration work in practice.