Nuclear power plants (NPPs) require accurate measurement of mass flow rates. Advanced flowmeters have been widely applied in several current industries; however, the operating environment in NPPs is especially harsh because of high temperature, high radiation, and extremely corrosive conditions. Several of the advanced reactor designs, such as liquid sodium pool reactors and integral small modular reactors, do not have conventional primary piping systems. These designs require an alternative method to accurately measure primary flow. Cross-correlation function (CCF) flow estimation can estimate the flow velocity indirectly without any specific instruments for flow measurement. The target flow rate is derived by the delay time between two sensors located near each other along the flow direction. Temperature sensors are a common choice for this function because they are reliable, economical, and widely used in various industries. The delay time is inferred by applying the CCF to the signals collected from two or more sensors. CCF flow estimation can be performed in any structure of the flow region, not limited to pipes. One challenge for the CCF flow estimation is that the accuracy of the flow measurement is mainly determined by the inherent local process variation, which is small compared to the uncorrelated noise. To differentiate the process variations from the uncorrelated noise, this paper demonstrates periodic fluid injection of a different temperature before the sensors to amplify common process variation. The feasibility and accuracy of this method have been investigated through a physical flow loop experiment designed to verify the CCF flow estimation using fluid injection. Several parameters must be selected when designing the fluid injection CCF measurement system, such as the distance between the fluid injection site and the sensors, the injection period, and the injection flow rate. A series of tests was conducted to investigate whether these parameters were related to the accuracy of the CCF flow estimation and to identify appropriate values for these parameters for different flow regimes. The results show that the fluid injection method improves the flow measurement performance, and the appropriate design of flow injection and measurement geometry produces better flow characterization performance over a range of flow rates.