A novel and simple assay system using a 96-well ELISA plate was established for measuring serum oxytocin in cynomolgus monkeys. This method omits the centrifuge for B/F separation because the second anti-rabbit IgG antibody-coated ELISA plate can easily separate the first anti-oxytocin rabbit antibody-bound radiolabeled oxytocin. Since this method has the advantage of omitting B/F separation, it becomes possible to measure a large number of samples with simple steps. In addition, accurate and reproducible results could be obtained by this method. The optimal reaction condition made it possible to measure more than 8 pg/ml of serum oxytocin. The changes of serum oxytocin level in relation to the first delivery was determined in a total of 11 female monkeys who were divided into two groups, infant-accepting mothers (4 monkeys) and infant-rejecting ones (7 monkeys). The serum oxytocin levels of pre-delivery (one to 4 days before delivery) and post-delivery (within 12 hr after delivery) in infant-accepting mothers were 33.6 +/- 4.57 and 43.5 +/- 16.4 pg/ml, respectively. Those in infant-rejecting mothers were 39.0 +/- 9.6 and 31.4 +/- 7.0 pg/ml. Two-way ANOVA (accepting/rejecting x pre/post) revealed a significant interaction of two factors (F (1, 9) = 5.39, p < 0.05). This result implies the possibility of a different pattern of oxytocin secretion between infant-accepting and infant-rejecting mothers during parturition.