This paper investigates the application of a threedimensional Bloom Filter (3DBF) to accomplish a secure and efficient accounts storage system by exploiting hashes of usernames and their corresponding passwords. We conducted numerical experiments and mathematical analysis to study the efficiency level of several 3DBF schemes. Our experimental results and analysis show that the level of occupancy for 3DBF is positively correlated to the value of its false positive rate, viz., if the occupancy level increases then so does the value of the false positive rate. Moreover, we also derive a formula for determining the minimum number of bits for storing some data in a 3DBF scheme given the value of its acceptable false positive rate and its occupancy level. We infer that the product of the dimensional parameter of a 3DBF scheme is inversely proportional to the false positive rate and occupancy level used in the scheme.