Two different models of 190 L (50 gallon) residential electric water heaters were tested to investigate problems that arise in determining their energy efficiencies due to the steep temperature gradient near the lower heating element. This gradient makes the determination of the average tank water temperature difficult because, in accordance with current test standards, temperature measurements are made at only six discrete locations within the water heater. Results show that errors in determining this average temperature can have significant effects on the Energy Factor by yielding inaccurate estimates of the stored energy within the tank. This estimate of the stored energy is part of the correction algorithm used to normalize the lab measured efficiency to the standard conditions that define the Energy Factor. The investigation’s findings suggest that efforts should be made to ensure that conditions at the start of a test are similar to those at the end. For the two water heaters tested here, this scenario was best achieved by starting with a 24 hour idling period.Units: Dual