A testing system for testing a touch screen graphical user interface is disclosed. The testing system includes a touch screen for displaying the graphical user interface under test and for receiving a first plurality of touch inputs from at least one user, a first module for recognizing the first plurality of touch inputs received by the touch screen as a second plurality of data structures, a second module for receiving the second plurality of data structures from the first module and constructing at least one test model, a third module for generating a third plurality of test cases based on the at least one test model constructed by the second module, and a fourth module for executing the third plurality of test cases generated by the third module.