Efficient bifunctional electrocatalyst for water splitting is essential for replacing fossil-fuel energy sources with clean energy-dense hydrogen fuel (142 MJ/kg). Efficient electrocatalyst can be obtained by either increasing active site density or specific activity on individual active sites. The active site densities can be increased through roughening the potential energy surface or exposing the facets which has higher active site densities. The specific activity can be increased through modulation of strain or charge densities on active sites which can be achieved through introduction of dopants, defects or stabilization of “non-native phases” that are all the other crystalline and amorphous states that differ in terms of discrete translational symmetry in the sub-surface region from the “native” phase (or bulk ground-state). While for a given composition, there is a unique native state for a given set of thermodynamic condition while, there can be many non-native structures having different bond-angles, bond-distances and surface atom densities from the native phase, leading to different electrocatalytic properties. In this context, polymorphic engineering via stabilizing ‘non-native phase’ offers a potential approach for improving both hydrogen evolution reaction (HER) and oxygen evolution reaction (OER) electrocatalysts and its activity. The beneficial effect of polymorphic engineering with regards to bifunctional electrochemical OER and HER is demonstrated by first principle calculation by taking CoSe2 as a model electrocatalyst which has marcasite (Space Group-58) and pyrite (Space Group-205) as the native (N) and non-native (NN) structures, respectively. The first principle computations predict pyrite (NN) structure of CoSe2 would have better electrochemical activity towards OER and HER than its marcasite (N) counterpart which is confirmed through experimental results in literature too. Though the co-ordination number of Co remains same in both the structures, the co-ordination symmetry surrounding Co atom varies. This results in differential charge distribution in constituting Co- and Se-atoms consequently resulting in variable density of state (DOS) near Fermi level (Figure 1) thereby affecting the binding energies (BE) of reaction intermediates of OER and HER. Pyrite (NN) phase of CoSe2 has a greater electron density near Fermi Level in comparison to its marcasite (N) counterpart due to differential co-ordination symmetry. A greater electron density near Fermi-level is indicative of lower work function and consequently lower polarization resistance during water splitting. A greater electron density near Fermi level is contributed by Co-3d orbitals which is the common active site for both OER and HER. The greater electron density and lower work function in Pyrite (NN) results in stronger metal-hydrogen BE (0.03 eV) resulting in lower overpotential of HER. Hydrogen adsorption on Se sites occurs only at higher HER overpotential due to weak Se-hydrogen BE (0.59 eV). This results in observation of twin Tafel slopes during HER on CoSe2 electrocatalyst as the potential determination step (PDS) switches from Volmer to Heyrovsky step with participation of Se during HER. The lower work function and higher electron density near Fermi level in Pyrite (NN) structure results in weaker metal-oxygen bond thereby promoting multi-electron OER activity. The OER intermediates (-OH, -O, -OOH) has a higher BE over Co- than Se-sites. The transformation of Oads à HOOads on Co-sites of CoSe2 (001) structure is the potential determination step with an onset potential of 1.66 V (vs RHE).The desorption of O2 from Se site is found to be the potential-determination-step (PDS) for OER (η=0.79 V). Furthermore, pristine CoSe2 acts as a precursor for OER which undergoes dissolution to form a surface Co-O structure which has a greater activity than pristine pyrite CoSe2 surfaces (η=0.31 V). This energetics is more favourable for pyrite (NN) structure than marcasite (N) structure for dissolution process to form surface Co-O structure due to stronger Co-Se bonds present in the latter case. Furthermore, point-defects which can aid both OER and HER, can be more easily formed in pyrite (NN) structure than marcasite (N) structure due to the aforementioned reason. The present study underlines the importance of stabilization of non-native structures which has a great potential to produce higher electrocatalytic activity thus providing greater options in search of better water splitting electrocatalyst. Figure 1