Effective green stormwater infrastructure (GSI) design requires comprehensive quantification of the volume of water that can be treated or removed over a given time period. It is recognized that evapotranspiration (ET) can be a substantial pathway for stormwater volume reduction in bioretention systems. However, measuring ET is often difficult and expensive, such as with lysimeters or a mass balance approach. This research focused on a new technique for quantifying ET in bioretention systems by exploring an approach using thermal imaging to calculate ET by measuring the flux of energy at the canopy surface. This thermal imaging approach was compared to ET measurements given by a traditional mass balance approach. The experimental setup had three benchtop scale vegetated lysimeters planted with Switchgrass. Time lapse thermal images of the Switchgrass plants were taken at 10 second intervals and paired with meteorological data. The data were used in an energy balance to estimate the mass of water lost from the lysimeter plant/soil system. That mass was compared to the change in weight measured by weighing the lysimeter before and after the data collection period. For comparison, reference ET was also calculated for the vegetated systems using three common reference ET equations. The uncalibrated energy balance equation developed here estimated an averaged ET over 12 data collection days within 1 mm of the mass balance measured ET. These findings demonstrate the feasibility of using a thermal image energy balance technique to estimate ET.