Multiple-stage tandem mass spectrometry and collision-induced dissociation were used to investigate loss of H2O or CH3OH from protonated versions of GGGX (where X = G, A, and V), GGGGG, and the methyl esters of these peptides. In addition, wavelength-selective infrared multiple photon dissociation was used to characterize the [M + H - H2O](+) product derived from protonated GGGG and the major MS3 fragment, [M + H - H2O - 29](+) of this peak. Consistent with the earlier work [Ballard, K. D.; Gaskell, S. J. J. Am. Soc. Mass Spectrom. 1993, 4, 477-481; Reid, G. E.; Simpson, R. J.; O'Hair, R. A. J. Int..J. Mass Spectrom. 1999, 190/191, 209-230000], CID experiments show that [M + H - H2O](+) is the dominant peak generated from both protonated GGGG and protonated GGGG-OMe. This strongly suggests that the loss of the H2O molecule occurs from a position other than the C-terminal free acid and that the product does not correspond to formation of the b(4) ion. Subsequent CID of [M + H - H2O](+) supports this proposal by resulting in a major product that is 29 mass units less than the precursor ion. This is consistent with loss of HN=CH2 rather than loss of carbon monoxide (28 mass units), which is characteristic of oxazolone-type b(n) ions. Comparison between experimental and theoretical infrared spectra ''or a group of possible structures confirms that the [M + H -H2O](+) peak is not a substituted oxazolone but instead suggests formation of an ion that features a five-membered ring along the peptide backbone, close to the ammo terminus. Additionally. transition structure calculations and comparison of theoretical and experimental spectra of the [M + H - H2O - 29](+) peak also support this proposal.