Serum rat PRL concentrations were compared using values determined by RIA and the Nb2 lymphoma cell bioassay (Nb2BA). Rat serum samples were obtained under different physiological conditions and after the administration of pharmacological agents known to affect PRL secretion. Of the treatments examined, estrogen, morphine bromocriptine and haloperidol significantly altered the relationship between Nb2BA and RIA estimates of PRL. The estrogen-induced increase in PRL levels of ovariectomized females and the proestrus surge of PRL in intact females led to higher bioassay than RIA estimates of PRL. Treatment with haloperidol, bromocriptine and morphine altered the relationship, favoring immunoreactive more than bioactive hormone, and reversing the pretreatment Nb2BA/RIA ratio. No discrepancies between estimates of PRL by the two assays were noted in untreated males, diestrous, estrous and ovariectomized females, or following ether or TRH administration. These results confirm previous observations of discrepancies between rat serum bioassay and RIA estimates, and the data suggest that different forms of prolactin are present in the circulation at different times.