Abstract: A method for measuring performance of test points having one or more instructions in a code sequence. A code sequence with two or more test points is repetitively executed, while measuring the execution time. The code sequence is modified by removing one test point. The modified code sequence is repetitively executed, again measuring the execution time. The modification and execution steps are repeated until no test points remain in the current code sequence. All execution and modification steps are repeated a defined number of times, saving the execution times to a data file. The minimum execution time for each code sequence from any run is assigned to that code sequence. The execution time of a particular test point is calculated by subtracting the assigned execution time for the code sequence after removing the particular test point from the assigned execution time for the code sequence be fore removing the particular test point.