Extracts of porcine hypothalamic fragments (HF) bring about the release of prolactin when injected into estrogen-progesterone pretreated male rats. To determine the extent to which this prolactin-releasing activity (PRA) is attributable to thyrotropin-releasing hormone (TRH) and/or vasopressin (VP), (both hormones capable of releasing prolactin in this preparation), PRA was assayed following destruction of TRH and VP by incubation in rat serum, and after separation on Sephadex G-25 columns. Acetic acid (2N) extracts of HF contain 22 to 27 ng TRH and 650 to 1000 ng of VP per HF as determined by immunoassay. Incubation for 1 h in fresh rat serum degraded 91 to 99% of both TRH and VP. PRA fell after incubation, but was still detectable, indicating residual activity that resisted degradation. Prolactin release responses to HF extracts and to TRH were log-dose dependent, but had different activity slopes. The minimal detected dose of TRH which released prolactin was 10 ng, while minimal effective doses of serum inactivated HF extract contained only 0.6 ng of TRH. Maximum effects with serum-inactivated HF extract were achieved with 2 HF equivalents containing 2.6 ng of TRH. More than 400 ng of TRH were required to give an equivalent PRA response. Sephadex G-25 chromatography of hypothalamic extracts using 2.0 N acetic acid separated a fraction which after treatment with serum to inactivate most TRH present caused marked prolactin release and contained only 0.7 ng of TRH and 0.3 ng of VP per dose. Evidence for a PIF was the demonstration that retarded fractions from the column significantly decreased plasma prolactin levels. The finding of PRA in hypothalamic extracts separate from both TRH and VP is evidence for the existence of a distinct prolactin-releasing factor.