Parameter determining argon phase Currently I am working a molecular simulation to determine phases of an argon NPT ensemble using Lennard Jones potential. Mainly I use the radial distribution function to determine solid, liquid, or vapor state by investigating the structure.
A problem comes in when I notice that there are some states which should be in vapor form that exhibit the radial distribution function of liquid (showing a widen peak). I surmise that there may be due to the high density gas especially above the critical point.
Is there any way to decisively differentiate the liquid and vapor phase?
 A: Not a complete answer, but hope it will help. Have you heard about SklogWiki? Among others it has a nice article about Lennard-Jones model.
The thing is that you are not dealing with argon, instead you are dealing with a model Lennard-Jones fluid. So it is wrong to suggest that phase transitions of the two perfectly coincide. Molecular potential assumed to be the true potential of argon is referred to as Aziz potential ("A highly accurate interatomic potential for argon", paywalled).
I would suggest you to compare your results with ones already known for LJ fluid, for example SklogWiki links to a work by JP Hansen, L Verlet "Phase Transitions of the Lennard-Jones System."
Apart from this I have a couple of remarks. You mention "states above critical point" --- there is no gas or liquid above the critical point, it is a state known as "supercritical fluid". It is normal you can't distinguish vapor from liquid there. The other thing I would be concerned of is the coexistence of vapor and liquid in the simulation. If it takes place everything is wrong. However I have never encountered this and I don't know how to diagnose such a nasty case.
