Testing should be stopped when
all the planned tests have been run
time has run out
all faults have been fixed correctly
it depends on the risks for the system being tested