Objectives The aims of this study were to describe the characteristics of hyponatremia in elderly users of antidepressants, to determine the prevalence and risk factors for hyponatremia, and to identify the underlying mechanisms. Study design Cross-sectional study (March 2007-April 2009) with prospectively collected data. Patients were older than 60 years, used antidepressants, and had a complete geriatric assessment. Main outcome measures Serum sodium and antidiuretic hormone levels, serum osmolality, urine sodium level, and urine osmolality were measured. The prevalence of hyponatremia (serum sodium <135 mM) as an adverse reaction to an antidepressant (AR-AD), defined with Naranjo's algorithm, was calculated. Hyponatremic patients were compared to normonatremic patients with regard to gender, age, weight, history of hyponatremia, hyponatremia-associated medications and disorders, and type and duration of antidepressant use. Results Of 358 eligible patients, 345 were included. The prevalence of hyponatremia as an AR-AD was 9.3%. Risk factors were a history of hyponatremia (adjusted OR 11.17, 95%CI 2.56-40.41), weight < 60 kg (adjusted OR 3.47, 95%CI 1.19-10.13), and psychosis (adjusted OR 3.62, 95%CI 1.12-11.73). Non-suppressed ADH was found in a minority of hyponatremic patients. Conclusions In elderly patients, the prevalence of hyponatremia as adverse reaction to all types of antidepressants was 9%. Patients with previous hyponatremia, weight <60 kg, and psychosis were at risk. Beside SIADH, the nephrogenic syndrome of inappropriate antidiuresis, in which ADH secretion was normal, is postulated as an underlying mechanism. This has consequences for treatment of antidepressant-induced hyponatremia with vasopressin receptor antagonists.