An online partitioning strategy is presented, in which test cases are selected based on feedback information collected during the testing process. The strategy differs from con- ventional approaches because the partitioning is performed online rather than off-line and because the partitioning is not based on program code or specifications. It can, therefore, be implemented in the absence of the source code or specification of the program under test. The cost-effectiveness of the proposed strategy has been empirically investigated with a set of subject programs, namely, SPACE, SED, GREP, and the Siemens Suite of Programs. The results demonstrate that the proposed strategy constantly achieves large savings in terms of the total number of test case executions needed to detect all faults.