Generating good quality test data is difficult and expensive, yet it is vital for software reliability. The TeTra project aims to remove barriers to automated generation of test data using a specifically adapted form of program transformation called `testability transformation'. This will require the formulation of new kinds of transformation which preserve properties relevant to test-coverage and the construction of novel algorithms based upon this formulation. TeTra will therefore extend research in both transformation and testing and will provide a link between the two areas of research activity (which have previously been considered to be disjoint). In order to achieve this, the TeTra team will develop theory, algorithms and tools for testability transformation and will evaluate these using case studies provided by the industrial collaborators. The principal deliverables will be: