Previous research on growth hormone-releasing factor analogues has used pituitary cell culture assay systems to evaluate in vitro their biological activity. However, binding assay systems in which receptor affinity and peptide stability can be assessed independently have been lacking so far. Since we have recently developed a sensitive GRF binding assay with [125I-Tyr10]hGRF(1-44)NH2, this method was applied to structure-affinity studies as a first step of screening GRF analogues. Acylation of the N-terminus of hGRF(1-29)NH2 generally decreased its affinity (relative affinity to hGRF(1-29)NH2 (RA), 26-85%). Replacement of the C-terminal carboxamide by a free carboxylic function decreased affinity likely by diminishing its proteolytic stability (RA, 57%). Removal of Tyr1, Ser9, Lys12, Val13, Gly15, Gln16, or Lys21 drastically decreased its affinity (RA, less than 3%). Multiple amino acid deletions in the segment 13-21 of hGRF(1-29)NH2 also led to a loss of affinity as did replacing segment 13-15, 16-18, or 19-21 by an octanoyl moiety (RA, less than 1%). Removal of Asn8, Gln24, Asp25, Ile26, Met27, and Ser28 or Arg29 had less effect on GRF receptor affinity (RA, 5-33%). Removal of Met27 or Ser28 only slightly affected hGRF(1-29)NH2 affinity (RA, 62-78%). Altogether, these results indicate that the amino acids contained in the segment 13-21 are more important than those of 24-29 to insure high affinity receptor binding or to maintain an optimal conformation to allow GRF binding.