The absolute bioavailability of an aqueous solution of 1-deamino-8-D-arginine vasopressin (dDAVP) from different regions of the gastrointestinal (GI) tract (stomach, duodenum, jejunum, ileum, colon, rectum) has been studied in 6 healthy, male volunteers aged 24 to 35 years, followed for 12 h after each drug administration. For i.v. administration the subjects received 4 micrograms dDAVP. For intestinal administration 400 micrograms dDAVP was directly applied to six distinct sites in the GI tract via two or four channel tubes with or without a distal occlusive balloon. Biological effects were assessed and plasma and urinary levels of dDAVP were measured using a specific, sensitive RIA. Urine osmolality remained elevated and diuresis decreased for 12 h following dDAVP administration irrespective of the site of application. After i.v. administration, the half-life of elimination of dDAVP was 60.0 min, plasma clearance 1.7 ml.min-1.kg-1, amount excreted in urine 2.0 micrograms and renal clearance was 0.8 ml.min-1.kg-1. The mean bioavailability (f) after gastric application was 0.19% (range 0.02-0.35%). f was 0.24% after duodenal application (range 0.04-0.62%), 0.19% after jejunal (range 0.01-0.41%), 0.03% after distal ileal (range 0.01-0.08%), 0.04% after proximal colonic (range 0.01-0.12%) and 0.04% after rectal (0.01-0.10%) application. The bioavailability was significantly higher in the three upper GI regions in comparison to the three lower regions. The bioavailability of dDAVP after gastric, duodenal and jejunal application was similar to that after swallowing a tablet in a previous study. Absorption from the ileum was lower than expected and no preferential site of absorption was found.