A radioimmunoassay to determine arginine-vasopressin concentration in rat plasma is described. This method was a modification from a previously described technique from which the incubation volume was reduced. It allows us to use only 300 microliter of plasma, 1/4 of the original volume, without any impairment to the sensitivity. The intra- and inter-assay coefficients of variation are 11% and 13% respectively. The accuracy ranges between 90 and 99%. To investigate the influence of sampling blood conditions on arginine-vasopressin levels, five groups of Sprague-Dawley rats are analyzed: Group I, decapitation; Group II, ether anaesthesia and cardiac puncture; Group III, ether anaesthesia and decapitation; Group IV, ketamine anaesthesia and cardiac puncture; Group V, ketamine anaesthesia and decapitation. Arginine-vasopressin levels in Group V (2.2 +/- 0.8 pg/ml) are significantly lower than those from the other four groups and similar to those obtained by other authors using a chronic catheter.