This paper reports on the results of a comparative experiment between two Bowen ratio measurement systems conducted at the Petawawa National Forestry Institute, Chalk River, Ontario, in 1985. Both systems interchange the positions of the psychrometers in the vertical. One, termed the pivot system, turns the psychrometers through 180° over a fixed separation distance, and inverts them in the process. It is also capable of orienting the psychrometers into the wind. The other, termed the elevator system, interchanges the psychrometers by moving them up and down two adjacent vertical tracks in a fixed orientation. The experiment lasted eight days, and the weather conditions varied from clear and dry to overcast. On the basis of hourly values of evapotranspiration, the agreement between the systems is within 10% for daytime values. When period totals are considered, this agreement improves to better than 5% for most days. There was no improvement shown by allowing the pivot system to orient into the wind. The elevator system is superior for two principal reasons. First, the separation distance between the psychrometers is easily varied to accommodate different sites. Second, during reversal there is no disturbance of the wet-bulb reservoirs which might lead to an oversupply of water to the sensors.