Software testing is an important method of assuring quality and enhancing reliability and path coverage is an essential criterion for testing adequacy.Since the rigid path cover testing is infeasible

we select and test some key paths

which outweigh others on affecting the overall quality of the system

to enhance test efficiency and profit.A general statistics-based automated tool aiming at object-oriented software systems was proposed.By static analysis and interpolation on code

this tool offers a diagnosis of statistical data derived from continual executions

which are driven by random test cases

for information of selecting key paths in correctness testing and reliability testing.Primary idea

corresponding algorithms of this tool

and some problems in implementation are exposed.