Essential fatty-acid-deficient rats were supplemented with 300 mg/d of pure fatty acid esters: oleate (O), linoleate (L), arachidonate (A), and columbinate (C) for 10 d. The 24-h urine collections from each animal, collected 3 d before supplementations and again the last 3 d of the 10-d supplementation period, were analyzed for volume, and by radioimmunoassay for arginine-vasopressin (AVP) and prostaglandin E (PGE). Linoleate and arachidonate supplements both decreased the initial high urinary AVP excretion, whereas it was further increased by the oleate supplement. There was no effect of columbinate supplementation on urinary AVP excretion. Urinary PGE excretion was increased ca. twofold by both linoleate and oleate supplements, increased ca. fivefold by arachidonate supplementation but was unaffected by columbinate supplementation. There was no effect of any of the supplemented fatty acids on urine output. Fatty acid analysis of total kidney lipids revealed a low percentage of 20:3(n-9) in the rats supplemented with (n-6) fatty acid. (L, A and C). The triene-tetraene ratio was 1.8 ± 0.6 (n = 6) in the kidneys of the oleate-supplemented rats. No relationship was found between urinary PGE excretion and the percentage of arachidonate or the ratio of 20:3 (n=9)/20:4(n-6) in total kidney lipids. It is suggested that increased urinary AVP excretion in EFA-deficient rats is mainly caused by a change in the renal excretatory mechanism of AVP rather than reflecting an increased plasma AVP concentration. Furthermore it is suggested that renal PGE synthesis in vivo is unaffected by high levels of 20:3(n-9) in kidney lipids.