It remains controversial whether two or three random estimations of prolactin (PRL) concentration obviate the need for special testing to rule out stress-related hyperprolactinemia. In order to clarify this issue, we measured PRL, cortisol, and growth hormone (GH) in serial blood samples obtained under resting conditions (HPR test) in 70 women who had had high PRL levels in two or more random blood samples. Twenty out of 70 women were found to have stress-related hyperprolactinemia, and 11 of the 20 would have been misdiagnosed by using only three random samples. Cortisol levels from the HPR test indicated stress-related pituitary adrenal reactivity in all groups, including the idiopathic (N = 30) and prolactinoma (N = 20) groups. However, PRL secretion in response to stress--a downward trend as a function of time--was evident only in the stress-related hyperprolactinemia group. These results suggest that a limited HPR test (ie, serial PRL determinations at 0, 30, and 60 min) is a valuable and simple measure to identify stress-related hyperprolactinemia in order to avoid diagnostic pitfalls and unnecessary treatment.