Trees provide a multitude of ecosystem services but are vulnerable to failure and limb loss under high winds. This is a natural process which initiates regeneration in forests but tree failures close to critical infrastructure networks lead to disruption to services and financial loss. Hence, network operators tend to apply the precautionary principle and remove all trees close to such infrastructure which leads to unnecessary loss of healthy trees, therefore, a more focussed approach is required. We introduce TREEFALL: an objective and scalable framework to assess tree failure risk. It uses novel approaches to quantify tree geometry, downscale wind parameters, simulate shielding by neighbouring trees and calculate wind-induced failure risk based on meteorological data for previous storms, scenarios or forecasts. Consequently, TREEFALL can identify individual trees which pose the greatest threat to infrastructure networks which can be targeted for field survey and management interventions where necessary. The model has broad potential for application to many different types of infrastructure networks and across the forest and environmental sciences.