In desorption electrospray ionisation (DESI) the interaction between the electrospray and the surface is key to two important analytical parameters, the spatial resolution and the sensitivity. We evaluate the effect of the electrospray solvent type, organic solvent fraction with water, analyte solubility and substrate wettability on DESI erosion diameter and material transferral into useful ion signal. To do this five amino acids, glycine, alanine, valine, leucine and phenylalanine are prepared as thin films on three substrates, UV/ozone treated glass, glass and polytetrafluoroethylene (PTFE). Four different solvents, acetonitrile (ACN), methanol (MeOH), ethanol (EtOH) and propan-2-ol (IPA), are used with organic solvent fractions with water varying from 0.1 to 1. These model systems allow the solubility or wettability to be kept constant as other parameters are varied. Additionally, comparison with electrospray ionisation (ESI) allows effects of ionisation efficiency to be determined. It is shown that the DESI efficiency is linearly dependent on the solubility (for these materials at least) and for analytes with solubilities below 1.5 g kg(-1), additional strategies may be required for DESI to be effective. We show that the DESI erosion diameter improves linearly with organic solvent fraction, with an organic solvent fraction of 0.9 instead of 0.5 leading to a 2 fold improvement. Furthermore, this leads to a 35 fold increase in DESI efficiency, defined as the molecular ion yield per unit area. It is shown that these improvements correlate with smaller droplet sizes rather than surface wetting or ionisation.