The general message is: whenever possible avoid volume relaxation with the default energy cutoff. Either increase the basis set by setting ENCUT manually in the INCAR file, or use method two suggested below, which avoids doing volume relaxations at all. If volume relaxations are the only possible and feasible option please use the following step by step procedure (which minimizes errors to a minimum):
The relaxation will give a structure which is correct to first order, the final error in the energy of step 3 is of second order (with respect to the structural errors). If you take the energy directly from the relaxation run, errors are usually significantly larger. Another important point is that the most accurate results for the relaxation will be obtained if the starting cell parameters are very close to the final cell parameters. If different runs yield different results, then the run which started from the configuration which was closest to the relaxed structure, is the most reliable one.
We strongly recommend to do any volume (and to lesser extend cell shape) relaxation with an increased basis set. ENCUT=1.3 default cutoff is reasonable accurate in most cases. PREC=High does also increase the energy cutoff by a factor 1.25. At an increased cutoff the Pulay stress correction are usually not required.
However, if the default cutoff is used for the relaxation, the PSTRESS line should be set in the INCAR file: Evaluate the Pulay stress along the guidelines given in the previous section and add an input-line to the INCAR file which reads (usually a negative number):