Quantum mechanics has led to the biggest technological breakthroughs of last century. However, many paradoxes still remain in the foundations of quantum theory. In this research we study one effect that leads to such paradoxes; quantum correlations. Quantum nonlocality refers to the set of quantum correlations that are not explainable using local models. It is quantum nonlocality that has troubled physicists since the inception of quantum theory nearly 100 years ago. However, these nonlocal quantum correlations offer a range of advantages in quantum information tasks. For example, quantum key distribution and quantum computing are two tasks that are powered by quantum correlations. For this reason a wide range of quantum correlations are studied experimentally in this research. These include: entanglement, EPR-steering, Bell nonlocality and bilocality. We study these correlations using the framework of quantum information, using single photons as our chosen quantum information carriers. The first quantum correlation tested in this research is EPR-steering. EPR-steering is the application of the EPR-paradox to a quantum information task. We implement three different experiments on EPR-steering. By studying the properties of the quantum information task we find that: EPR-steering is more robust to depolarisation noise than Bell-nonlocality; EPR-steering is absolutely loss tolerant; and tests of EPR-steering are easier to carry out, than tests of Bell nonlocality, but harder to implement compared to tests of non-separability. As a result of these three findings we experimentally show that EPR-steering can be demonstrated on Belllocal states, we close the detection loophole in a photonic test of quantum nonlocality, and we discover and implement maximally parsimonious tests of non-separability, EPR-steering and Bell nonlocality.