Chemical data obtained through LA-ICP-MS, EPMA and XRF analysis have been used to investigate the potential for provenancing archaeological.artefacts. By combining chemical analysis with multivariate statistical analysis (pCA and cluster analysis) a more robust and objective method of classification and provenancing can be obtained. Modem glass bottles were used as an analogue for Islamic Bahraini glass samples, to test the application ofXRF analysis. For the modem glass samples which were homogeneous, with the chemistry being tightly controlled during manufacture and were free from alteration by weathering, an accurate method of provenancing was obtained. The colour of the glass was found to be a confounding factor. By restricting statistical analysis to a single colour of sample, a wide range of bottles from around the world were accurately grouped into the region where they were manufactured. Applying the same methods to the Islamic Bahraini glass it was found that the amount of weathering was a major confounding factor, as was the colour of the sample. When performing PCA and cluster analysis it was found that samples clustered together depending on their degree of weathering or how well weathering products had been removed prior to analysis. The size of sample for XRF was found to have an effect on the accuracy of the analysis, caused by x-ray attenuation within the sample and therefore requiring a ZAP like correction to be made to major and minor elemental data. No effect was seen when performing PCA and cluster analysis. Variations in the amounts of MgO and K20 present indicate that plant ashes were used as the soda source and not natron. There may be slight differences in the plants or the parts of plants used in manufacture. Ah03 contents also indicate differences in the silica source but neither of these appear to be related to the period of time or location where the samples were found. It is possible that raw glass was brought to Bahrain for re-working or that finished items were traded from elsewhere as no evidence was found in the samples for primary glass production (frit). The analysis of carnelian beads from India and Mali by LA-ICP-MS was found to be difficult. Although the heterogeneity within a carnelian bead was found to be much less than that seen in samples of agate, the chemical variations within in a sample was still large and would outweigh any differences between sources.