Database represents an essential part of software applications. Many organizations use database as a repository for large amount of current and historical information. With this context testing database applications is a key issue that deserves attention. SQL Exception handling mechanism can increase the reliability of the system and improve the robustness of the software. But the exception handling code that is used to respond to exceptional conditions tends to be the source of the systems failure. It is difficult to test the exception handling by traditional methods. This paper presents a new technique that combines mutation testing and global optimization based search algorithm to test exceptions code in Jordan University Hospital (JUH) database application. Thus, using mutation testing to speed the raising of exception and global optimization technique in order to automatically generate test cases, we used fitness function depends on range of data related to each query. We try to achieve the coverage of three types of PL/SQL exceptions, which are No\_Data\_Found (NDF), Too\_Many\_Rows (TMR) and Others exceptions. The results show that TMR exception is not always covered this due to existence of primary key in the query, also uncovered status appear in nested exceptions.