Evapotranspiration (ET) is a key component of the energy and water balances in permafrost tundra, establishing hydrological conditions for the next year and controlling several aspects of the carbon cycle. Both the energy balance and hydrological conditions of the landscape surface are important drivers of how Arctic climate change will impact landscape processes, including the carbon feedback. The accurate measurement of evapotranspiration within an energy balance context therefore provides crucial information on ecosystem functioning and raises our predictive capacity for estimating the impact of climate change. In this study we report field measurements from 13 summers (2002-14) using the eddy covariance method in a lowland ice-wedge polygon landscape within Russia’s Lena River Delta. These time-series are gap-filled and extrapolated with both statistical and process-based models to generate estimates of growing season ET. We find that interannual differences – including two August periods with high ET and two with low ET – are locally driven more by changes in air temperature and vapor pressure deficit (VPD) than in land surface characteristics or radiation. Except for periods of high VPD, aerodynamic resistance was greater than canopy surface resistance. We explore predictive relationships between various land surface indicators (e.g., NDVI, LAI, LST, Growing season length) derived from remote sensing products (MODIS) to quantify local mechanisms necessary for upscaling to the Delta region. Nighttime land surface temperature (MODIS) is found to be a strong predictor of evaporative flux at weekly to monthly time scales. Contrary to expectations resulting from climate change studies, we do not see evidence of a sustained interannual trend in ET or sensible heat flux. We conclude with implications for the local energy balance and responses to changes in sea ice extent and a warming climate.