In biological sample microanalysis, a mass-normalisation method is commonly used as a quantitative index of elemental concentrations determined by particle-induced X-ray emission (PIXE). The organic mass can either be determined using particle backscattering spectrometry (BS) or scanning transmission ion microscopy (STIM). However, the accuracy of quantitative microanalysis in samples such as cultured cells is affected by beam-induced loss of organic mass during analysis. The aim of this paper is to compare mass measurements determined by particle BS or by STIM. In order to calibrate STIM and BS analyses, we measured by both techniques the thickness of standard foils of polycarbonate (3 and 6 mum), Mylar (R) (4 mum), Kapton (R) (7.5 mum) and Nylon (R) (15 mum), as well as biological samples of mono-layered cultured cells. Non-damaging STIM analysis of samples before PIXE irradiation is certainly one of the most accurate ways to determine the sample mass, however, this requires strong experimental handling. On the other hand, BS performed simultaneously to PIXE is the simplest method to determine the local mass in polymer foils, but appears less accurate in the case of cultured cells. (C) 2001 Elsevier Science B.V. All rights reserved.