SummaryRats were subjected to the following procedures: No treatment, Stressor (10% NaCl i.p.), Warfarin for 7 days, Stressor followed by Warfarin; and groups were sacrificed at intervals for assessment of spontaneous hemorrhage and of adrenal ascorbic acid concentration. There was no hemorrhage in the no treatment and stressor groups; some hemorrhage in the warfarin group; profound hemorrhage with Warfarin + Stressor. The adrenal ascorbic acid concentration was found to be lower, 8 h and again 5 days after stress, and remained lower in the warfarin + stress animals. Warfarin had no effect on adrenal ascorbic acid level.In another series of experiments in which the stress consisted of an electric current to the cage floor for 6 sec over 15 min, rats were sacrificed daily for determination of serum corticosterone concentration and occurrence of spontaneous hemorrhage. There was a statistically significant increase of serum corticosterone concentration with stress, warfarin and combined warfarin and stress treatments (P< 0.001 for all three variables). There was a significant correlation (r = 0.96 and 0.89, P< 0.01) for serum corticosterone concentration with hemorrhage score and incidence of hemorrhage in stressed rats receiving warfarin, but not in those receiving only warfarin. The results indicate an activation, rather than an exhaustion, of the pituitary-adrenal axis during the combined action of anticoagulant and stress, which results in the development of spontaneous hemorrhage.