Middle East Current Psychiatry 2011, 18:29–36 Introduction For many years, researchers have suggested that abnormalities in circadian rhythms may underlie the development of mood disorders, such as major depression. The objectives behind this study were to investigate whether cortisol circadian rhythm could be considered as a biomarker for major depression, to identify whether different circadian rhythm alterations in patients with major depression is interrelated with syndrome presentation (either with or without psychotic features) and symptom profile (typical and atypical depressive symptoms) and to investigate whether there are alterations in the cortisol circadian rhythm and whether it could be a possible endophenotype among first-degree relatives of patients. Methods Fourteen patients, aged between 25 and 50 years, in a major depressive episode or in a depressed phase of bipolar disorder, six controls and six first-degree relatives were screened using psychometric questionnaires, and a blood sample was withdrawn from consenting participants at four time points during the day, for the assessment of plasma cortisol secretion. The patients were diagnosed by a clinical psychiatric interview and a semistructured psychiatric interview based on the Diagnostic and Statistical Manual of Mental Disorders, 4th Edition, Text Revision criteria for diagnosis, and a psychometric assessment was made. Serum cortisol levels were assayed using a validated commercial kit. Results There was a statistically significant difference between serum cortisol levels of the three studied groups at four consecutive sampling points over 24 h. The elevated plasma cortisol levels of patients correlated positively with phenotypic features, such as suicidality (P=0.029), positive family history of psychiatric disorder (P=0.024), chronotype of the patients (P=0.017) and their scores on Beck Depression Inventory (P=0.029) and Hamilton Depression Rating Scale (P=0.05). In addition, there is no statistically significant difference between patients and their first-degree relatives on the scores of Bipolar Affective Disorder Dimension Scale. Conclusion The increase in cortisol secretion and shift in its circadian phase and chronotypes of depressed patients and their first-degree relatives suggest that cortisol level is probably a state marker in mood disorder. In addition, different phenomenology and symptomatology of depression was closely interrelated to cortisol levels. Further genetic studies are recommended to establish its predictive significance.