Two complete and two partial structure-activity relationship scans of the active fragment of human growth hormone-releasing hormone, [Nle27]-hGHRH(1-29)-NH2, have identified potent agonists in vitro. Single-point replacement of each amino acid by alanine led to the identification of [Ala8]-, [Ala9]-, [Ala15]- (Felix et al. Peptides 1986 1986, 481), [Ala22]-, and [Ala28, Nle27]-hGHRH(1-29)-NH2 as being 2-6 times more potent than hGHRH(1-40)-OH (standard) in vitro. Nearly complete loss of potency was seen for [Ala1], [Ala3], [Ala5], [Ala6], [Ala10], [Ala11], [Ala13], [Ala14], and [Ala23], whereas [Ala16], [Ala18], [Ala24], [Ala25], [Ala26], and [Ala29] yielded equipotent analogues and [Ala7], [Ala12], [Ala17], [Ala20], [Ala21], and [Ala27] gave weak agonists with potencies 15-40% that of the standard. The multiple-alanine-substituted peptides [MeTyr1,Ala15,22,Nle27]-hGHRH(1-29)-NH2 (29) and [MeTyr1,Ala8,9,15,22,28,Nle 27]-hGHRH(1-29)-NH2 (30) released growth hormone 26 and 11 times, respectively, more effectively than the standard in vitro. Individual substitution of the nine most potent peptides identified from the Ala series with the helix promoter alpha-aminoisobutyric acid (Aib) produced similar results, except for [Aib8] (doubling vs [Ala8]), [Aib9] (having vs [Ala9]), and [Aib15] (10-fold decrease vs [Ala15]). A series of cyclic analogues was synthesized having the general formula cyclo(25-29)[MeTyr1,-Ala15,Xaa25,Nle27,Yaa29+ ++]-GHRH(1-29)-NH2, where Xaa and Yaa represent the bridgehead residues of a side-chain cystine or [i-(i + 4)] lactam ring. The ring size, bridgehead amino acid chirality, and side-chain amide bond location were varied in this partial series in an attempt to maximize potency. Application of lactam constraints in the C-terminus of GHRH(1-29)-NH2 identified cyclo(25-29)[MeTyr1,Ala15,DAsp25,Nle27,Orn29+ ++]-hGHRH(1-29)-NH2 (46) as containing the optimum bridging element (19-membered ring) in this region of the molecule. This analogue (46) was 17 times more potent than the standard. Equally effective was an [i-(i + 3)] constraint yielding the 18-membered ring cyclo(25-28)[MeTyr1,Ala15,Glu25,Nle,27Lys28]- hGHRH-(1-29)-NH2 (51) which was 14 times more potent than the standard. A complete [i-(i + 3)] scan of cyclo(i,i + 3)[MeTyr1,Ala15,Glui,Lys(i + 3),Nle27]-hGHRH(1-29)-NH2 was then produced in order to test the effects of a Glu-to-Lys lactam bridge at all points in the peptide. Of the 26 analogues in the series, 11 had diminished potencies of less than 10% that of the agonist standard, 4 were weak agonists (15-40% relative potency), and 4 analogues were equipotent to the standard. The 7 most potent analogues ranged in potency from 3 to 14 times greater than that of the standard and contained the [i-(i + 3)] cycles between residues 4-7, 5-8, 9-12, 16-19, 21-24, 22-25, and 25-28. The combined results from these systematic studies allowed for an analysis of structural features in the native peptide that are important for receptor activation. Reinforcement of the characteristics of amphiphilicity, helicity, and peptide dipolar effects, using recognized medicinal chemistry approaches including introduction of conformational constraints, has resulted in several potent GHRH analogues.