Consideration of the stress ellipsoid shows that the Principal Horizontal Stress (PHS) direction may not be a determinable or particularly significant stress direction for some kinds of faults. Principal stress magnitudes, as well as principal stress directions, can play a role in fixing the orientation of PHS. The usefulness of a method for PHS location recently proposed by Lensen is questioned on this basis and also because considerable errors may arise from its application to faults for which the PHS direction is determinable. Fault dip and the angle between a fault and the greatest principal stress direction are considered important factors given insufficient attention in Lensen's analysis. An alternative method for location of stress directions is discussed briefly.