Quantitative Evaluation of Materials by Scanning Electron Microscopy (QEMSCAN®) analysis of thick sections of lava and tephra samples from the 2021 Tajogaite eruption from La Palma, Canary Islands. QEMSCAN® data were processed to produce a phase map and an olivine composition map per sample, in which each distinct phase/composition was assigned an 8-bit pixel intensity value. Olivine compositions were divided into 4 equal ranges of forsterite content (using the Fe-Mg ratio in the raw data) which were later calibrated using EPMA data. Pixels that did not fit the strict raw data requirements needed to be assigned a mineral phase were assigned to “Undifferentiated groundmass” , and instead reflect pixels containing glass or a combination of minerals ± glass below the pixel resolution. Isolated pixels of chemically complex phases such as amphibole may be the result of partial- or sub-volume interaction (i.e. between clinopyroxene, Ti-magnetite, plagioclase) and not a true representation of phase abundance. Isolated and small clusters of QEMSCAN® pixels of amphibole were shown by BSE data to be particularly prone to these effects, and so were added to the groundmass. Individual phase maps were extracted, labelled and analysed in Python using the pandas, numpy and skimage libraries. Pixel connectivity was determined by any of 4 neighbours, as opposed to 8 neighbours (any adjacent pixel including on the corners), which was more accurate in delineating groundmass crystals from each other. We chose an area of 10,000 microns2 to reflect the minimum size of phenocrysts, and assigned every region smaller than this limit to the groundmass. Phases that contributed =5% phenocrysts by area (normalised to solid rock area) in at least one sample were regarded as “major”, while the rest termed “minor”. Full details of the QEMSCAN® methodology is found in the accompanying Supplementary Data1_QEMSCAN spreadsheet.