Properly quantifying evapotranspiration (ET) is a critical step in determining water and energy balances, especially in Arctic landscapes where spatial and temporal heterogeneity in soil water content and inundation is pronounced. Although the eddy covariance technique has gained popularity as an approach for estimating ET at aggregate scales, obtaining ET estimates at finer spatial scales remains problematic. Thus, ET is poorly estimated for highly variable tundra landscapes, despite the importance of this process for parameterization and validation of models. To overcome this methodological limitation, we developed an approach to measure diurnal ET by modifying a LI-8100A (LI-COR, Lincoln, NE, USA), a chamber-based instrument typically used for measuring soil CO₂ fluxes. To enable the use of the LI-8100A for ET determinations, a calibration method was designed and implemented through laboratory and independent field measurements in Arctic and semi-arid locations. Once calibrated, the instrument was deployed June–September 2013 for diel measurements of ET on the Arctic coastal plain near Barrow, Alaska, USA. We validated the system by comparison to four adjacent plots measured by a LI-6400-09 soil CO₂ flux system that was also calibrated to calculate water vapour flux. In conclusion, we determined that with calibration, the LI-8100A can make long-term, high-frequency measurements of ET, even in low flux, continuous-permafrost landscapes. This technique provides an opportunity to assess fine-scale ET and its topographic controls across low-centre and high-centre polygons and to rigorously compare such measurements with aggregate fluxes obtained with eddy covariance.