1. The main aim of this study was to identify putative beta-bends and the role of the N- and C-terminus in the CGRP receptor antagonist halpha CGRP8-37, which was measured against halpha CGRP inhibition of twitch responses in the rat prostatic vas deferens. 2. With a bend-biasing residue (proline) at position 16 in halpha CGRP8-37 (10(-5) M) an inactive compound was produced, while alanine at the same position retained antagonist activity (apparent pKB 5.6+/-0.1 at 10(-5) M). Proline at position 19 within halpha CGRP8-37 (10(-5) M) was an antagonist (apparent pKB 5.8+/-0.1). 3. Incorporation of a bend-forcing structure (beta-turn dipeptide or BTD) at either positions 19,20 or 33,34 in halpha CGRP8-37 (10(-5) M) antagonized halpha CGRP responses (apparent pKB 6.0+/-0.1 and 6.1+/-0.1, respectively). Replacement by BTD at both positions 19,20 and 33,34 within halpha CGRP8-37 competitively antagonized responses to halpha CGRP (pA2 6.2; Schild plot slope 1.0+/-0.1). 4. Halpha CGRP8-37 analogues (10(-5) M), substituted at the N-terminus by either glycine8, or des-NH2 valine8 or proline8 were all antagonists against halpha CGRP (apparent pKB 6.1+/-0.1, 6.5+/-0.1 and 6.1+/-0.1, respectively), while halpha CGRP8-37 (10(-5) M) substituted in three places by proline8 and glutamic acid10,14 was inactive. 5. Replacement of the C-terminus by alanine amide37 in halpha CGRP8-37 (10(-5) M) failed to antagonize halpha CGRP responses. 6. Peptidase inhibitors did not alter either the agonist potency of halpha CGRP or the antagonist affinities of halpha CGRP8-37 BTD19,20 and 33,34 and halpha CGRP8-37 Gly8 (against halpha CGRP responses). 7. In conclusion, two beta-bends at positions 18-21 and 32-35 are compatible with high affinity by BTD and is the first approach of modelling the bioactive structure of halpha CGRP8-37. Further, the N-terminus of halpha CGRP8-37 is not essential for antagonism, while the C-terminus interacts directly with CGRP receptor binding sites of the rat vas deferens.