BACKGROUND: Secondary data from different policy sectors can provide unique insights into the social, environmental, economic and political determinants of health. This is especially pertinent in the context of whole-systems approaches to healthy public policy, which typically combine cross-sectoral collaboration with the application of theoretical insights from systems science. However, the sharing and linkage of data between different sectors are still relatively rare. Previous research has documented the perspectives of researchers and members of the public on data sharing, especially healthcare data, but has not engaged with relevant policy and practice decision-makers. AIM: We sought to work collaboratively with decision-makers relevant to healthy public policy and practice in Scotland to identify practical ways that cross-sectoral data sharing and linkage could be used to best effect to improve health and reduce health inequalities. METHODS: We facilitated three sequential stakeholder workshops with 20 participants from local and central government, public health teams, Health and Social Care Partnerships, the third sector, organisations which support data-intensive research and public representatives from across Scotland. Workshops were informed by two scoping reviews (carried out in June 2021) and three case studies of existing cross-sectoral linkage projects. Workshop activities included brainstorming of factors that would help participants make better decisions in their current role; reflective questions on lessons learnt from the case studies; and identifying and prioritising recommendations for change. Findings were synthesised using thematic analysis. SETTING AND SCOPE: Scotland; public and third sector data. RESULTS: Based on the workshops, and supported by the reviews and case studies, we created a visual representation of the use of evidence, and secondary data in particular, in decision-making for healthy public policy and practice. This covered three key overarching themes: differing understandings of evidence; diverse functions of evidence; and factors affecting use (such as technical, political and institutional, workforce and governance). Building on this, workshop participants identified six guiding principles for cross-sectoral data sharing and linkage: it should be pragmatic; participatory; ambitious; fair; iterative; with holistic and proportionate governance. Participants proposed 21 practical actions to this end, including: a strategic approach to identifying and sharing key data sets; streamlining governance processes (e.g. through standardised data sharing agreements; central data repositories; and a focus on reusable data resources) and building workforce capacity. To make these possible, participants identified a need for strong political and organisational leadership as well as a transparent and inclusive public conversation. LIMITATIONS: Participation from some stakeholders was limited by workload pressures associated with the COVID-19 pandemic. No consensus was reached on the impact, effort, and/or timing of some recommendations. Findings were closely informed by the Scottish context but are nonetheless likely to be relevant to other jurisdictions. CONCLUSIONS: There is broad consensus among key stakeholders that linked cross-sectoral data can be used far more extensively for public health decision-making than it is at present. No single change will lead to improved use of such data: a range of technical, organisational and political constraints must be addressed. FUNDING: This article presents independent research funded by the National Institute for Health and Care Research (NIHR) Public Health Research programme as award number NIHR133585.