The current test methods that evaluate the liquid water resistance performance of a water resistive barrier (WRB) have various drawbacks. A Hydrostatic Pressure Test Apparatus (HPTA) provided by SMT Research Ltd. was used to develop a unique test method to detect liquid water penetration through WRBs. The sensor used in HPTA has 42 channels and each channel reads an individual electric resistance. Lower electric resistance indicates more liquid water penetrates the sensor, vice versa. The measurement system of the device was studied and tested. The results showed the measurement system works properly to detect amount change of the liquid water applied to the sensor while there is no proportional relationship between the amount of liquid water and electrical resistance. The change between two adjacent electrical readings was evaluated by a factor called Normalized Difference Vegetation Index (NDMI). The NDMI factor was determined to be a better measurement than the value of electrical resistance. Preliminary tests using HPTA with Membrane D and Membrane F were done and the apparatus was found to have water leaking through the joint. The clamping method of HPTA was then modified to solve the problem and further WRBs were tested using the modified device. Six WRB materials, Membrane A, Membrane B, Membrane C, Membrane D, Membrane E and Membrane F were tested and 83 tests were conducted in total. A colormap method was used to evaluate the water resistance ability of each material. Among all materials, Membrane E had the highest water resistive duration of long and a water resistance score (WRS) of 337.14. Membrane F had a lower duration of short and a WRS of 78.75. Membrane C and Membrane D have water resistance durations between very short and short and their WRS were 36.5 and 82.71. Membrane A and Membrane B had the lowest water resistance duration of very short and their WRS were 15.15 and 23.19. Plots of Log R vs. Time and NDMI vs. Time of the edge sensors and center sensors were plotted in MATLAB. The water movement behaviours in the two groups of sensors varied by materials.