The association between selective serotonin reuptake inhibitors (SSRis) and hyponatraemia has been well documented, the elderly appearing to be at greatest risk. An analysis of data of hyponatraemia in the elderly using SSRis from all published cases and from the Committee on Safety of Medicines found that the mean time to detection was about 3 weeks after commencing SSRis. A wide range of time to detection (1-253 days) and non-specific symptoms suggest hyponatraemia is detected by chance rather than being specifically looked for. This is probably a sporadic, idiosyncratic phenomenon that is not dose related as A VP function determined by serum and urine concentrations was found to be normal in six elderly patients using sertraline. In the elderly there are physiological changes, a high prevalence of medical illnesses and concomitant drug use which may precipitate hyponatraemia. Together with a risk of altered water regulation in psychiatric illness this may account for the particular susceptibility of this group to hyponatraemia whilst using SSRis. AIMS & HYPOTHESIS: This dissertation will explore the physiology of Arginine V asopressin and how changes in this system along with other physiological changes in the elderly make the elderly susceptible to hyponatraemia. This problem will then be explored in the context of elderly people with depression using SSRis which are known to cause hyponatraemia. In the first part of the research section the aims are to report the published cases of hyponatraemia occurring whilst using SSRis from the United Kingdom and specifically focus on cases in people 60 years and older. Secondly to re-analyse all case reports in the literature looking only at this population. The third aim was to investigate whether dysregulation of vasopressin function in the elderly using SSRis is a sporadic or usual phenomenon. The null hypothesis is that A VP function is not disturbed by SSRis.