Major Depression does not always remit even with appropriate treatment, and unremitting or difficult-to-treat depression is thought to contribute to the large disease burden caused by depression. Difficult-to-treat depression is an overarching term that can be used to describe depression which is chronic, unremitting or treatment-resistant. The concept of difficult-to-treat depression and the terms used to describe it have not been well validated or universally adopted for use in research and clinical practice. Accordingly, the overarching aim of this thesis was to investigate difficult-to-treat depression, including its conceptualisation and its correlates, with a particular focus on chronic and treatment-resistant depression (TRD). Three methodological approaches were employed to examine this complex phenomenon: 1) a systematic review of the literature; 2) analysis of epidemiological data; and 3) analysis of clinical data. found to have a lifetime prevalence of 4.6% (95% CI: 3.9 -5.3%) and was found to be present in 29.4% (95% CI: 25.6 -33.3%) of community-residing individuals with a lifetime depressive disorder. Higher rates of psychiatric co-morbidity (OR = 1.42; 95% CI = 1.26-1.61), older current age (OR = 1.04; 95% CI = 1.02-1.05), a younger age of onset (OR = .97; 95% CI = .95-.98) and more frequent episodes of depression (OR = 1.75; 95% CI = 1.07-2.86) were significant correlates of chronic depression. Differences in health service utilisation associated with chronic depression were assessed to determine whether chronic depression treated in the tertiary care sector was likely to represent TRD. Survey respondents with chronic depression who were treated in tertiary care settings had more complex presentations and many of the clinical features previously associated with TRD, including higher levels of medical and psychiatric co-morbidity, greater traumatic load, higher levels of disability, greater symptom severity and greater risk of attempting suicide. iv The findings from the NSMHWB suggested that those whose chronic depression is treated in tertiary care settings may be less responsive to treatment, thus indicating likely TRD. In order to assess the degree of TRD in patients seen in tertiary care settings a sample of depressed inpatients (N = 70) was recruited. The majority of these depressed inpatients had a chronic illness trajectory (N = 64; 91.4%) and had a moderate to high level of TRD as determined by the five existing staging models of TRD. Each of the five staging models of TRD was highly correlated with the other four models, suggesting a substantial degree of agreement between models on their ratings of TRD. An omnibus measure of TRD (M = 0; SD = 4.5) was created by combining the five TRD models into one composite index. Using this composite index as the outcome variable, the following covariates were found to be associated with higher levels of TRD: higher prevalence of suicide attempts (β = 1.71, t(11) = 3.62, p < .001), older current age (β = 14, t(11) = 2.92, p < .005), earlier a...