Summary In the heavy-oil Atlanta field, well-test operations were planned initially with a multienergy gamma ray/Venturi multiphase flowmeter as a contingency. During the well-test operations, the foaming propensity of the produced oil made it almost impossible to distinguish an interface between liquid and gas in the separator or tank, rendering the standard equipment useless. Only in the second test were some tank measurements available. The measurement principle used in this multiphase flowmeter, however, is not sensitive to foam or emulsion because the meter responds to the atomic-level composition of the different components of the mixture independently of their arrangement, providing accurate flow-rate measurements. Therefore, it became the main flowmetering device for the operator. We detail the behavior of the multiphase flowmeter during the well tests in the Atlanta field and the challenges caused by this highly viscous oil. In the case of standard oils, the measurements are not sensitive to the viscosity value. However, in the case of low Reynolds numbers, the viscosity value can dramatically affect the results. In this case, some solvents were injected during the well test to reduce the liquid viscosity, and we present the consequences for the multiphase flowmeter. Because it was not possible to measure the viscosity directly at the wellsite, we highlight the work flow that was followed to determine the proper viscosity value to use. The viscosity number was validated by comparing the multiphase-flowmeter measurements to the gas/oil ratio measured in the laboratory on downhole samples from the same reservoir, to the water cut measured manually during the operations and to the tank total-liquid-rate measurements. Finally, we present some recommendations on the use of multiphase flowmeters for heavy-oil well tests.