1. Eight normal subjects underwent water loading alone and water loading plus 40 mg of frusemide IV, fluid intake ad libitum alone and fluid intake ad libitum plus frusemide, plus each of the preceding after pretreatment with indomethacin. 2. After frusemide administration, increases in urinary sodium excretion paralleled increases in urinary volume, and urinary prostaglandin E 2 (PGE 2 ) excretion correlated closely with sodium excretion ( y = 1.03 x −0.28; r = 0.940; P <.0001). 3. In the absence of the diuretic, urinary volume varied over a wide range with little change in sodium excretion. Again, urinary PGE 2 excretion correlated with urinary sodium excretion ( y = 0.12 x + 0.05; r = 0.789; P <.002). However, the correlation differed markedly from that observed in the studies with frusemide. 4. Expressing urinary PGE 2 excretion as a function of urinary volume for all of the studies resulted in a highly significant correlation ( y = 10.7 x −0.70; r = 0.975; P <.0001). 5. Multiple and stepwise regression analyses assessing the correlation of urinary PGE 2 excretion with urinary flow rate and with indices of function of various nephron segments indicate that the correlation with urinary PGE 2 could be predominantly accounted for by urinary volume. 6. We conclude that in the condition of this study in man, urinary PGE 2 excretion is a correlate of urinary volume.