In this thesis I propose a new way of understanding depressive illness as not exclusively harmful, but as related to particular, empirically evidenced, epistemic and pragmatic benefits for the subject, alongside the associated costs. For each of the benefits considered, I provide and concisely analyse the empirical evidence both in its favour and against it, suggest ways in which these benefits could apply in the circumstances presented, discuss some outstanding problems for that application as stated, and describe potential implications. The issues discussed involve topics in cognitive psychology as well as in ethics, which I regard as complementary to the debate. I take the view that depression related phenomena such as low mood, low self-esteem, and mood-congruent delusions offer potential benefits to a subject under certain circumstances, but not under others. In discussing the hypotheses presented, I argue for a more complex view of depression, which includes its potential epistemic and pragmatic benefits as well as its costs and I propose possible directions for future research.