Metabolic fingerprinting is increasingly employed in microbial and plant metabolomics. Identification and evaluation of analytical factors that influence mass spectra produced with automated electrospray time of flight mass spectrometry to support metabolic fingerprinting are described. Instrument resolution of 4000 (FWHM) at mass 200 Da provided detection of ions of the same nominal mass but different monoisotopic masses. Complex mass spectra were obtained from polar extracts of tomato fruit in positive and negative ion mode. These spectra consist of metabolite ions (molecular, adduct and fragment) and those derived from the extraction medium, largely in the form of [M+H], [M-H], [M+Na], [M+K], [2M+H], [M+Cl] and [2M-H]. Ionisation suppression reduced sensitivity, although its effect was consistent for a wide range of metabolite concentrations. Variability in ion signal intensity was lower in analytical (2.2-30.1%) compared to biological (within fruit 9.6-27.6%; between-fruit 13.2-34.4%) replicates. The method is applicable to high throughput metabolic fingerprinting and, with accurate mass measurements, is able to provide reductions in data complexity and preliminary identification of metabolites.