A method can be used for phase selection using multi-terminal measurements of a transmission line of a power system. The method includes obtaining, using measurements from a plurality of terminals of the transmission line, delta differential current values ΔIdiffA(t), ΔIdiffB(t), ΔIdiffC(t), at time t, between pre-fault differential current measurements and post-fault differential current measurements from all three phases (A, B, C) of the transmission line. The method also includes determining at least one of the phases to be faulty when the absolute value of the corresponding delta differential current value is larger than k times the minimum of all the absolute values IΔIdiffA(t)|, |ΔIdiffB(t)|, |ΔIdiffC(t)| of the delta differential current values, where k>1 is a scale factor.