Several studies indicate that the disappearance rate of ACTH in different species is very rapid. The conditions under which these estimates were obtained, however, can be regarded in most cases as nonphysiological. The present studies were designed to evaluate the half-life of endogenous ACTH in rats under physiological conditions, i.e. awaken freely moving animals that were bled continuously through an indwelling iv cannula. Blood sampling at 1-min intervals allowed the detailed characterization of the ACTH pulsatile pattern. Analysis of each secretory episode indicated that many of the pulses could be used for half-life determinations, provided that the down phase of the pulse consisted of three or more points decaying in a straight line. After logarithmic transformation of the values, the declining slopes were analyzed by linear regression to determine which slopes showed a significant negative correlation. From those values the half-life estimates, in minutes, were calculated. Two groups of animals were used for these studies. In one group, blood was removed continuously without replacement. In these animals, ACTH levels rose progressively due to the hemorrhage effect, although secretion continued to be pulsatile. In a second group, blood was replaced at the same rate of withdrawal through a femoral vein, and ACTH levels remained constant throughout the experiment. The results indicate that the average ACTH half-life was 5.94 and 7.06 min for the groups without and with blood replacement, respectively. The differences between the two groups were not significant. Analysis of the frequency distribution of the calculated half-lives indicates that two populations can be separated, a fast (decays less than 8 min) and a slow one (decays greater than 8 min). Less than half of the secretory episodes showed a decay according to the half-life. This observation indicates that some secretory episodes consist of a burst of secretory activity followed by a period during which secretion is essentially turned off. Under these conditions, decays correspond to the half-life of the hormone. Other episodes show protracted secretory activity, as indicated by the slower, nonlinear decay. These observations suggest that half-life of ACTH may vary from one episode to another, and that the different types of secretory episodes recorded may reflect the multifactorial control of ACTH secretion by the hypothalamus.