Monitoring the residues of test coverage in actual use or during beta testing activities can validate any assumptions of their impact on quality. But this is unlikely to be accepted by customers unless the performance impact is small. A previous work presented a proof-of-concept tool written for Java programs to demonstrate that the performance overhead of running instrumented executables is low if instrumentation is removed from already executed basic blocks. It is however not clear whether and to what extent this result depends on the high run-time overhead of the Java platform. It is also an open question whether residual coverage monitoring would have a significant advantage relative to complete coverage monitoring if the complete monitoring were implemented using the best known monitoring techniques. This work answers the above questions by using EEL. an tool for instrumenting compiled languages, a quick profiler. QPT2 and standard SPEC CINT95 benchmarks.