Moisture damage is one of the major causes of premature failure in asphalt pavements, and it also accelerates the severity of other distresses. To date, no moisture test has been widely accepted that is reliable and practical in predicting the field moisture performance of the asphalt mix during the design stage. One reason is because the sample conditioning methods cannot represent the field conditions, resulting in inconsistent results with the field performance of some mixtures. Taken into account this concern, this paper investigates how different testing conditions, including sample preparation, moisture saturation, and loading methods, can affect the results of laboratory moisture tests. In conclusion, it is found that the degree of vacuum pressure for achieving moisture saturation and air voids distribution has a significant impact on the moisture testing results. Multiple freeze–thaw cycles have a limited effect on the variation of mechanical performance (i.e., compressive dynamic modulus). If one or several freeze–thaw cycles are to be used in a moisture test, the effect of aging should be considered. It is recommended that a sample without coring and cutting should be used for a moisture test as the coring and cutting process is found to change the air voids distribution, i.e., the percent of connected air voids, thus making the sample not representative to the field condition. Finally, the moisture test results are more sensitive under tension mode than under compression mode.