Transpiration is an integral part of the earth system, not only because plant water use is the dominant path by which water flows from the soil to the atmosphere, but also because water loss from leaves is intrinsically connected to CO2 uptake and provides a key link between global carbon and water cycles. Transpiration, while well studied at the scale of plants and leaves, remains difficult to quantify at the ecosystem, regional, and global scales. Differentiating between water vapor which has passed through plants, and is thus biologically controlled, and water vapor which is evaporated from surfaces without active control from plants through stomatal regulation is extremely challenging at the ecosystem scale with heterogeneous landscapes containing diverse plant species accessing soil water reserves at varying depths. One potential solution to characterizing transpiration on broader scales is to use eddy covariance to estimate ecosystem transpiration. Eddy covariance has been widely utilized to measure water, carbon, and energy fluxes, with synthesis initiative such as FLUXNET collating hundreds of sites around the world. However, methods for partitioning the total ecosystem water flux (evapotranspiration, ET) measured by eddy covariance systems into the individual components, i.e. transpiration (T) and abiotic evaporation (E), are needed. The work presented here demonstrates the viability and utility of ecosystem scale estimates of transpiration from eddy covariance datasets via data driven methodologies. Identifying key strengths and uncertainties in the method, such as the uncertainty in the magnitude of transpiration but strength in spatial and temporal patterns, better outlines future directions.