Recent developments in laser ablation inductively coupled plasma-mass spectrometry (LA-ICP-MS) have enabled improvements in spatial resolution and analytical detection limits. Here, the analysis of individual glass shards from tephra deposits using a 193 nm Excimer laser (producing ablation craters as small as 4 μm diameter), coupled to a magnetic sector ICP-MS, is described. Analyses of individual glass shards with crater diameters of 20 μm and 10 μm is essentially routine, and when element fractionation is corrected for, good accuracy is achieved. Analytical precision is good, being around ±15–30% at 1 ppm and around ±2–3% at 500 ppm from 10 μm diameter ablation craters, and lower limits of detection (LLD) are <1 ppm for most elements from 10 μm craters, when 25–28 trace elements are determined in a ∼20 s analysis. Neither 44Ca nor 43Ca can be used reliably as the internal standard for the analysis of rhyolites from 10 μm ablation craters, because CaO is close to the lower limit of quantitation (LLQ); thus 29Si must be used, although either could be used in the analysis of basaltic glasses at 10 μm. With analyses at 6 μm or 4 μm, many trace elements in rhyolites (e.g. Zr, Ba, LREE, Y, Rb, U, Th) remain above the LLQ, but at this resolution, only Si can be used as an internal standard for glass analysis. Element fractionation is an issue for all analyses <20 μm in diameter, resulting from the formation of a thin melt film on the ablation crater walls. This melt film becomes an increasingly larger proportion of the volume of ablated material as crater diameters become smaller, because the surface area/volume ratio increases. Element retention or volatility from this melt film appears to cause much of the fractionation. For larger craters (20 μm) this fractionation appears to affect all compositions similarly. For many elements determined from smaller craters (10 μm–4 μm) there is a systematic variation in the degree of fractionation with the glass (and thus melt film) composition. This relates to a change in the degree of polymerisation of the glass, with, for example, the REE being the most fractionated in polymerised rhyolitic samples at the smallest crater diameters (4 μm). This systematic behaviour, however, offers some hope for the analysis of a selection of abundant trace elements in individual shards of glass using ablation craters of 6 μm and 4 μm in diameter.