Low permeability dome rocks may contribute to conduit overpressure development in volcanic systems, indirectly abetting explosive activity. The permeability of dome-forming rocks is primarily controlled by the volume, type (vesicles and/or microcracks), and connectivity of the void space present. Here we investigate the permeability-porosity relationship of dome-forming rocks and pumice clasts from Merapi’s 1888 to 2013 eruptions and assess their possible role in eruptive processes, with particular emphasis on the 2010 paroxysmal eruption. Rocks are divided into three simple field classifications common to all eruptions: Type 1 samples have low bulk density and are pumiceous in texture; Type 2 samples, ubiquitous to the 2010 eruption, are dark grey to black in hand sample and vary greatly in vesicularity; and Type 3 samples are weakly vesicular, light grey in hand sample, and are the only samples that contain cristobalite. Type 2 and Type 3 rocks are present in all eruptions and their permeability and porosity data define similar power law relationships, whereas data for Type 1 samples are clearly discontinuous from these trends. A compilation of permeability and porosity data for andesites and basaltic andesites with published values highlights two microstructural transitions that exert control on permeability, confirmed by modified Bayesian Information Criterion (BIC) analysis. Permeability is microcrack- and diktytaxitic-controlled at connected porosities, φc, < 10.5 vol.%; vesicle- and microcrack-controlled at 10.5 < φc < 31 vol.%; and likely vesicle-controlled for φc > 31 vol.%. Type 3 basaltic andesites, the least permeable of the measured samples and therefore the most likely to have originated in the uppermost low-permeability dome, are identified as relicts of terminal domes (the last dome extruded prior to quiescence). Cristobalite commonly found in the voids of Type 3 blocks may not contribute significantly to the reduction of the permeability of these samples, mainly because it is associated with an extensive microporous, diktytaxitic texture. Indeed, the low permeability of these rocks is more likely associated with their lower fracture density. We propose that diktytaxitic textures may arise from late-stage gas filter pressing of a silica-rich melt phase, which leaves behind a microlite-supported groundmass and cristobalite in neighbouring vesicles. Due to the ubiquity of the Type 3 rocks in all Merapi eruptions, we do not invoke the emplacement of a low-permeability cap as having favoured a particularly high pressurization and subsequent high explosivity of the 2010 eruption. The debate as to the reasons for the highly explosive 2010 eruption rages on.