Landscape-scale restoration planning requires cost estimates that reflect different site conditions and project goals. Estimates that assume a homogenous landscape are insufficient for project budgeting and optimizing scenarios. Currently, few resources and little guidance exist for those seeking restoration cost data. We developed a method to generate cost estimates for prairie and wet meadow restoration in the Upper Midwest United States. Using scenarios to provide hypothetical site-and project-specific detail, we surveyed restoration practitioners about their preferred actions and associated costs for 16 restoration transitions from common start states, reflecting vegetative cover, and end states, reflecting restoration targets and project goals. We estimated average seed cost for each end state using example seed mixes and pricing information from regional vendors. Baseline hydrologic restoration costs for wet meadows were estimated via expert consultation. We summed average costs for seed and individual restoration actions to produce total estimated costs for each transition. As expected, average restoration transition costs varied widely, with costs ranging more than four-fold. Differences were driven primarily by seed cost and intensity of vegetation management. Practitioners also varied considerably in both their reported sequence of actions and estimated costs for each action; individual transition costs ranged by as much as 25-fold. Within-practitioner variability of this magnitude presents serious challenges for restoration planning at multiple scales. Scenarios are an effective tool for generating coarse-resolution restoration cost estimates that reflect important site and project differences; however incentivizing practitioners to provide cost data and understanding sources of variability in their estimates remain important challenges.