To test the hypothesis that renal PGs may mediate the renal functional defect of potassium depletion, rabbits were placed on normal potassium and potassium-deficient diets for 7 weeks, and measurements were made of urinary PGE2 excretion; renal cortical, medullary, and papillary PGE2 content; and in vitro, de novo PGE2 biosynthesis. In the sixth week maximal urinary osmolality declined significantly, from 1118 +/- 44 mOsm/kg of H2O in controls to 666 +/- 25 in potassium-deficient animals, accompanied by a corresponding decrease in urinary excretion of potassium from 12.2 +/- 0.7 mEq/24 hr to 0.93 +/- 0.1 and in muscle potassium content from 39.0 +/- 1.0 mEq/100 gm to 22.3 +/- 1.9. By 3 weeks urinary excretion of PGE2 was significantly lower in hypokalemic animals than in controls. Renal cortical, medullary, and papillary PGE2 tissue content decreased significantly, from 0.061 +/- 0.01 microgram/gm to 0.022 +/- 0.005, 0.73 +/- 0.09 microgram/gm to 0.267 +/- 0.036, and 8.6 +/- 1.04 microgram/gm to 4.6 +/- 1.4, respectively. In vitro PGE2 biosynthesis by cortical, medullary, and papillary slices from normal animals was 0.102 +/- 0.01, 9.55 +/- 1.7, and 25.5 +/- 2.8 microgram/gm/30 min, whereas corresponding values from hypokalemic rabbits were 0.04 +/- 0.006, 3.9 +/- 0.747, and 14.0 +/- 1.25 microgram/gm/30 min, respectively. These results indicate that renal slices from hypokalemic rabbits synthesize much less PGE2 in vitro than do normal controls. The date do not support the hypothesis that enhanced PGE2 synthesis mediates the renal concentrating defect of chronic potassium deficiency but suggest that decreased renal PGE2 production may underlie the impairment in renal hemodynamics known to exist in this condition.