We seek to identify the major differences between two types of peak-fitting programs for alpha-particle spectrometry. One function is the well-known convolution of a Gaussian function with two (usually) left-handed exponential functions. The other function is contained in an analysis program described by Koskelo et al. (Koskelo, M. J., Burnett, W. C. and Cable, P. H. (1996) An advanced analysis program for alpha-particle spectrometry. Radioactivity Radiochem. 7(1), 18) and has one tailing parameter which can be adjusted to generate a modified Gaussian which is skewed towards lower energies. Recently an inter-laboratory exercise (Euromet 325) has provided some data which can be used to compare the two fitting functions. We present some results using the SRIM program (Ziegler, J. F., Monte-Carlo Program TRIM contained in Program suite SRIM version 96.xxx, 1996) and discuss what might be the underlying physical phenomena described by the fitting functions.