Password recovery tools are often used in digital forensic investigations to obtain the passwords that are used by suspects to encrypt potential evidentiary data. This paper presents a new method for deterministically generating and efficiently storing password recovery tables. The method, which involves the virtual expansion of rainbow tables, achieves improvements of 16.92% to 28.15% in the password recovery success rate compared with the original rainbow table method. Experimental results indicate that the improvements are achieved with the same computational complexity and storage requirements as the original rainbow table method.