There is an ongoing debate regarding the robustness and credibility of published scientific research. In this article, the authors argue that these issues stem from two broad causal mechanisms: the cognitive biases of researchers and the incentive structures within which researchers operate. To address these issues, the UK Reproducibility Network (UKRN) was founded and is introduced in this paper. The UKRN is supporting several initiatives at various levels across the UK research system and is investigating the factors that contribute to robust research, promoting training activities, disseminating best practice, and working with stakeholders (like researchers, institutions, funders, publishers, and others) to ensure coordination of efforts across the sector.