There is growing need to quantify and communicate how land use and management activities influence soil organic carbon (SOC) at scales relevant to, and in the tangible control of landowners and forest managers. The continued proliferation of publications and growth of datasets, data synthesis and meta-analysis approaches allows the application of powerful tools to such questions at ever finer scales. In this analysis, we combined a literature review and effect-size meta-analysis with two large, independent, observational databases to assess how land use and management impact SOC stocks, primarily with regards to forest land uses. We performed this work for the (Great Lakes) Lake States, which comprise 6% of the land area, but 7% of the forest and 9% of the forest SOC in the U.S., as the second in a series of ecoregional SOC assessments. Most importantly, our analysis indicates that natural factors, such as soil texture and parent material, exert more control over SOC stocks than land use or management. With that for context, our analysis also indicates which natural factors most influence management impacts on SOC storage. We report an overall trend of significantly diminished topsoil SOC stocks with harvesting, consistent across all three datasets, while also demonstrating how certain sites and soils diverge from this pattern, including some that show opposite trends. Impacts of fire grossly mirror those of harvesting, with declines near the top of the profile, but potential gains at depth and no net change when considering the whole profile. Land use changes showing significant SOC impacts are limited to reforestation on barren mining substrates (large and variable gains) and conversion of native forest to cultivation (losses). We describe patterns within the observational data that reveal the physical basis for preferential land use, e.g., cultivation of soils with the most favorable physical properties, and forest plantation establishment on the most marginal soils, and use these patterns to identify management opportunities and considerations. We also qualify our results with ratings of confidence, based on their degree of support across approaches, and offer concise, defensible tactics for adapting management operations to site-specific criteria and SOC vulnerability. The three distinct data sources used in this analysis are archived as part of this dataset.