Agro-ecosystems are the dominant land-use type on Earth, covering more than a third of ice-free land surface. Agricultural practices have influenced the Earth's climate system by significantly altering the biogeophysical and biogeochemical properties from hyper-local to global scales. While past work has focused largely on characterizing the effects of net land cover changes, the magnitude and nature of gross transitions and agricultural management practices on climate remains highly uncertain. To address this issue, a new set of global gridded land-use forcing datasets (LUH2) have been developed in a standard format required by climate models for CMIP6. For the first time, this dataset includes information on key agricultural management practices including crop rotations. Crop rotations describe the practice of growing crops on the same land in sequential seasons and are essential to agronomic management as they influence key ecosystem services such as crop yields, water quality, carbon and nutrient cycling, pest and disease control. Here, we present a data-driven approach to infer crop rotations based on crop specific land cover data, derived from moderate resolution satellite imagery and created at an annual time-step for the continental United States. Our approach compresses the more than 100,000 unique crop rotations prevalent in the United States from 2013 - 2015 to about 200 representative crop rotations that account for nearly 80% of the spatio-temporal variability. Further simplification is achieved by mapping individual crops to crop functional types, which identify crops based on their photosynthetic pathways (C3/C4), life strategy (annual/perennial) and whether they are N-fixing or not. The resulting matrix of annual transitions between crop functional types averages ~41,000 km²/yr for rotations between C3 and C4 annual crops, and ~140,000 km²/yr between C3 N-fixing and C4 annual crops. The crop rotation matrix is combined with information on other land-use states to compute annual changes between these states, thereby producing a detailed land-use transition information that can help close regional and global carbon budgets. We also validate the quality of the crop rotations identified in our product in countries with agronomic practices different from the United States.