This paper presents a technique that leverages an existing regression test-selection algorithm to compute accurate, updated coverage data on a version of the software, Pi+1, without rerunning any test cases that do not execute the changes from the previous version of the software, Pi, to