Source determination of use-related residues on prehistoric stone tools is especially challenging, due to issues related to preservation, contamination and the contribution of residues from multiple sources. To increase confidence in this process, an analytical workflow was developed to include: (1) a sampling strategy that retains spatial information of the recovered residues and enables monitoring of environmental contamination; and (2) a sensitive and selective gas chromatography-tandem mass spectrometry (GC-MS/MS) procedure to quantify non-volatile low molecular weight lipids on stone artefacts. This workflow was applied to 14 stone artefacts excavated from deposits at Liang Bua, a limestone cave on the Indonesian island of Flores. These artefacts range in age between ∼14 000 and 1000 years old, and were preliminarily classified as either potentially showing traces of use (n = 7) or not (n = 7) using low magnification microscopy. Residues were sampled by direct solvent extraction off the surface of the artefacts. The aliquots were spiked with internal standards and derivatised. The trimethylsilyl derivatives of 40 saturated fatty acids, sterols, di- and tri-terpenoids and their analogues were quantified using optimised multiple reaction monitoring (MRM) transitions. Six of the potentially used artefacts contained sterols, phytosterols and terpenoids, either individually or in combination, whereas none of these compounds was commonly found on the seven artefacts preliminarily classified as unused. This suggests that these six artefacts may have been used as implements to process resources, and provides scope for further investigation. This workflow can also be adapted for the analysis of other archaeological objects.