The diurnal response of ACTH release to intravenously administered arginine vasopressin was tested in normal volunteers given consecutively moderate doses of vasopressin every 15 min (0.1, 0.3, 1.0, and 3.0 IU) at 2200 h and again at 0700 h (PM/AM). This protocol was repeated 4 wk later with the times reversed (AM/PM). A dose-related increase in ACTH secretion was observed in all subjects. When the AM response of the AM/PM protocol was compared with the PM response of the PM/AM protocol, the release of ACTH was greater in the morning (P less than 0.05) as evaluated by the following criteria: peak value of ACTH (129.9 +/- 30.4 pg/ml in the AM vs. 57.1 +/- 20.2 in the PM); area under the curve (689 in the AM vs. 259 in the PM); and, sensitivity of the ACTH dose-response curve (first significant increase in ACTH with 1 IU of vasopressin in the AM but not significant even after 3 IU in the PM). In addition, when the AM vasopressin testing followed a previous evening stimulation (PM/AM protocol), there was a blunted ACTH response compared with the AM/PM protocol. Corticotropin-releasing factor (CRF) is probably the major ACTH secretagogue, but since vasopressin acts synergistically with CRF to produce an augmented release of ACTH, we suggest that the ACTH response to administered vasopressin depends upon the ambient endogenous level of CRF. We interpret our data and published data that CRF produces a lesser release of ACTH in the AM as follows: in the morning endogenous CRF is high and administered CRF produces little further release of ACTH, but administered vasopressin acting synergistically with high endogenous CRF causes a greater release of ACTH; conversely, in the evening endogenous CRF is low and administered CRF causes a greater release of ACTH, but vasopressin (a weak secretagogue by itself) gives a low ACTH response. We conclude that vasopressin stimulation of ACTH secretion can be used as an in vivo bioassay of endogenous CRF, and that there is a diurnal rhythm of CRF in hypophyseal portal blood.