A key challenge in conservation is the efficient allocation of limited resources to maximise benefits for biodiversity. Decision-support tools that account for landscape heterogeneity are needed to identify spatially-explicit actions that will achieve the greatest biodiversity benefits with available resources. We developed a raster-based, landscape-scale, spatial conservation action planning tool (SCAP) that offers significant advances for prioritising local and regional scale conservation actions in heterogenous landscapes. The SCAP tool was developed for the state of Victoria, Australia, to integrate heterogeneity of landscapes, species distributions, threats, and management costs and benefits across the state. We used empirical data to derive current and pre-European settlement distributions for 4400 native terrestrial species, and developed spatially explicit models of 19 threats to biodiversity. We coupled structured expert-elicitation techniques with machine learning to map the expected benefits to species, and the implementation costs, of 17 management actions – alone and in combination. We then ranked location-specific actions by their cost-effective contribution to an overall objective of minimizing the risk of species loss in Victoria over the next 50 years, using a modified implementation of the Zonation conservation planning framework. The SCAP tool provides decision makers with a transparent decision-support tool for identifying the cost-effective management actions at scales relevant to management.