This study investigated the repeatability and influences of the time of the day (TOD) and horn status on cortisol responses to ACTH administration in heifers. Sixty-four heifers were subjected to three ACTH challenges. The first challenge (C1) took place at the age of 2 mo. Balanced for peak cortisol responses at C1, the heifers were assigned to one of two rearing conditions: horned (H+) or disbudded (H−). At the age of 15 months, the second (C2) and third (C3) challenges took place, 7 d apart from each other at the same TOD. For cortisol analysis, saliva was sampled in 30-min intervals from 30 min before to 150 min after each ACTH injection. The area under the curve (AUC) of cortisol was calculated with respect to the ground (AUCG) and to the increase (AUCI). Between C2 and C3, AUC values did not differ (P > 0.10), intra-class correlation coefficients (ICCs) indicated poor repeatability (AUCG: ICC = 0.24 and AUCI: ICC = 0.26) and no correlations were found. The TOD had no effect on AUC values in C2 (P > 0.1), while in C3, they were greater in the morning than in the afternoon (for both AUCG and AUCI, P < 0.05). The H+ and H− heifers showed similar cortisol responses in C3, but in C2, horned heifers had greater AUC levels (P < 0.05). From C2 to C3, AUC values increased and decreased for heifers tested in the morning and afternoon, respectively. This was more pronounced in H+ than in H− heifers (interaction effect P <0.05). The results indicate poor to lacking repeatability for ACTH challenges performed within the same physiological state. While TOD and horn status partly contributed to the cortisol responses’ variance, the poor repeatability critically questions the use of repeated ACTH challenges for stress research in cattle.