As a first step in an extensive project planned to determine serum PRL levels in response to oral metoclopramide in women with a diverse gyneco-obstetric history, it was decided to study 51 clinically healthy nulliparous women, aged 15.8 to 48.2 years, with history of regular menses at least one year before the study (except the three postmenopausal women), with no regular drug ingestion during the last six months. Women were studied on days 18 to 22 of menstrual period, after a 30 minute rest on basal conditions (3 samples) at 60, 90, and 120 minutes after a single 10 mg. oral dose of metoclopramide. Duplicate PRL determinations were performed in all samples and progesterone(P) only in a pool of the three basal samples by radioimmunoanalysis. All women had serum P levels > or = 4.0 ng/ml. A significant linear positive correlation (r > or = 0.6795, p < 0.001) was observed between chronologic age (CA) and serum PRL levels, regardless the way they were expressed. Considering the individual responses it was decided to divide the group according to CA and it was observed that serum PRL levels--expressed in any form were always significantly greater in women aged > 25 years (Group 2) in contrast with women aged < or = 25 years (Group 1). Since differences were evident, percentiles 3, 50 and 97 for serum PRL levels were calculated during each test time for both groups.(ABSTRACT TRUNCATED AT 250 WORDS)