We study the efficiency of DUNE, a next generation long baseline oscillation experiment to resolve two major unknowns in neutrino oscillation physics. These are, octant of $\theta_{23}$ (i.e. if $\theta_{23}$ is $< 45^\circ$ or $>45^\circ$) and Dirac CP phase $\delta_{CP}$. We mainly focus on the role of antineutrinos when they travel 1300 km baseline of DUNE. We observe that for DUNE, the antineutrino runs help to remove parameter degeneracies even in the parameter space where the antineutrino probability suffers from various degeneracies. We study these points in detail and find that, due to enhanced matter effect longer baseline experiments create an increased tension between the neutrino and the antineutrino probabilities which helps to increase total sensitivity in case of combined runs. We also find that, antineutrino run increases overall CP sensitivity due to its ability to abolish octant-$\delta_{CP}$ degeneracy.