Two C-terminal and one N-terminal peptide fragments derived from corticotropin-releasing factor (CRF-1–41), CRF-28–41, CRF-34–41, and CRF-1–8, and the CRF receptor antagonist α-helical CRF-9–41 (αhCRF) were evaluated for their behavioral, endocrine and autonomic nervous effects in rats. To this purpose, three different approaches were used. First, rats were tested in a passive avoidance behavioral (PAB) task after intracerebroventricular (i.c.v.) injection the peptides at five different doses. I.c.v. CRF-1–41 was found to attenuate PAB. The effects of CRF-1–41 on PAB were completely antagonized by i.c.v. pretreatment with αhCRF. When given as a sole treatment, the antagonist produced a bimodal effect on PAB. At low doses, αhCRF tended to facilitate PAB, while high doses of the antagonist significantly attenuated PAB. Both CRF fragments showed behavioral effects similar to CRF-1–41. Of all peptides tested, CRF-34–41 was found to most attenuate PAB at both retention tests. In a second experiment, the behaviorally most potent fragment CRF-34–41 was compared to CRF-1–41 for its adrenocorticotropic activity after iv injection in pentobarbital-anesthetized rats. Treatment with CRF-1–41, in a dose-related fashion, produced a significant rise in plasma ACTH, whereas CRF-34–41 was without effect. Finally, the effect of i.c.v. injected CRF-1–41 and CRF fragments on heart rate (HR) and gross activity was measured in conscious rats in their home cages during a 60-min period, using a wireless telemetry system. Concomitantly, the occurrence of grooming behavior was recorded. During the first 10 min after i.c.v. treatment, of all peptides tested, CRF-34–41 produced the most marked increase in HR, which remained significant only during the first 30 min of recording. CRF-28–41 induced a non significant transient tachycardia. The parent molecule CRF-1–41 also induced an immediate, significant tachycardia but this effect lasted longer than 60 min. The N-terminal CRF-1–8 remained without effect. No significant effects on gross activity were observed with the two short C-terminal peptides, whereas i.c.v. injected CRF-1–41 induced excessive grooming behavior. A significant grooming response was also recorded in rats given CRF-34–41 fragment, but not in those treated with CRF-28–41 or CRF-1–8. The observed structure-related dissociation of endocrine, autonomic and behavioral effects suggests the possible existence of peptide fragments, derived from CRF-1–41 in the central nervous system, with distinct biological activity, acting on different receptors.