The effects of K depletion on urine concentrating ability, renal PG[prostaglandin]E2 excretion, and ADH [antidiuretic hormone] release were studied in 28 female dogs made K depleted by oral K-exalate and a K-free diet. After K depletion was established (serum K 2.9 .+-. 0.1 meq/l), urine volume increased from control measurements, 596.4 .+-. 34.0 to 1201.5 .+-. 96.9 ml/24 h (P < 0.001); urine PGE2 excretion increased, 985.4 .+-. 91.1 to 2122.0 .+-. 328.5 ng/24 h, (P < 0.001); and Umax decreased, 2006 .+-. 74.0 to 1186 .+-. 71.9 mOsm/kg H2O (P < 0.001). Indomethacin (5 mg/kg per day, s.c.) administered on 3 consecutive days after K depletion had been established, resulted in no significant improvement in Umax, 1186.8 .+-. 71.9 to 1341.8 .+-. 105.6 mOsm/kg H2O. Release of ADH from the neurohypophysis was evaluated by measuring plasma ADH during graded increases in serum tonicity with i.v. hypertonic saline before and after K depletion. Although ADH increased with increasing serum tonicity during both control and K depletion periods, there was a blunting of ADH release during K depletion. The regression coefficient of plasma ADH and serum tonicity was significantly lower during K depletion, 0.24, than in the control period, 0.65 (P < 0.01). After 3 days of indomethacin (5 mg/kg per day, s.c.) ADH release from the neurohypophysis in response to graded increases in serum tonicity was partially normalized in the K-depleted animals without changes in serum K (regression coefficient, 0.53). K depletion in dogs therefore leads to an increase in urine volume and an increase in renal PGE synthesis associated with a decrease in Umax. The increase in PGE2 synthesis is not responsible for the defect in Umax, since it is not corrected with indomethacin. The release of ADH in response to raising serum tonicity is blunted during K depletion, which is partially corrected by indomethacin. These data suggest an inhibiting role for PGE2 in the release of ADH from the neurohypophysis during K depletion.