Measures aimed at conservation or restoration of ecosystems are often seen as net‐cost projects by governments and businesses because they are based on incomplete and often faulty cost‐benefit analyses. After screening over 200 studies, we examined the costs (94 studies) and benefits (225 studies) of ecosystem restoration projects that had sufficient reliable data in 9 different biomes ranging from coral reefs to tropical forests. Costs included capital investment and maintenance of the restoration project, and benefits were based on the monetary value of the total bundle of ecosystem services provided by the restored ecosystem. Assuming restoration is always imperfect and benefits attain only 75% of the maximum value of the reference systems over 20 years, we calculated the net present value at the social discount rates of 2% and 8%. We also conducted 2 threshold cum sensitivity analyses. Benefit‐cost ratios ranged from about 0.05:1 (coral reefs and coastal systems, worst‐case scenario) to as much as 35:1 (grasslands, best‐case scenario). Our results provide only partial estimates of benefits at one point in time and reflect the lower limit of the welfare benefits of ecosystem restoration because both scarcity of and demand for ecosystem services is increasing and new benefits of natural ecosystems and biological diversity are being discovered. Nonetheless, when accounting for even the incomplete range of known benefits through the use of static estimates that fail to capture rising values, the majority of the restoration projects we analyzed provided net benefits and should be considered not only as profitable but also as high‐yielding investments.