The Social Amplification of Risk Framework (SARF) is often used as a conceptual tool for studying diverse risk perceptions associated with environmental hazards. While widely applied, it has been criticised for implying that it is possible to define a benchmark ‘real’ risk that is determined by experts and around which public risk perceptions can subsequently become amplified. It has been argued that this objectification of risk is particularly problematic when there are high levels of scientific uncertainty and a lack of expert consensus about the nature of a risk and its impacts. In order to explore this further, this paper examines how ‘experts’ – defined in this case as scientists, policy makers, outbreak managers and key stakeholders – construct and assemble their understanding of the risks associated with two invasive tree pest and disease outbreaks in the UK, ash dieback and oak processionary moth. Through semi-structured interviews with experts in each of the case study outbreaks, the paper aims to better understand the nature of information sources drawn on to construct perceptions of tree health risks, especially when uncertainty is prevalent. A key conclusion is that risk assessment is a socially-mediated, relational and incremental process with experts drawing on a range of official, anecdotal and experiential sources of information, as well as reference to past events in order to assemble the risk case. Aligned with this, experts make attributions about public concern, especially when the evidence base is incomplete and there is a need to justify policy and management actions and safeguard reputation.