Abstract: Magma viscosity controls the style of a volcanic eruption and thus its hazard potential, and is in turn controlled by magma temperature, water content, and crystallinity. However, magma viscosity can only be measured during or after an eruption. Identification of precursors that allow estimates of magma viscosity would greatly improve volcano forecasting. A synthesis of case studies of high-frequency volcanic microearthquakes (volcano-tectonic earthquakes, or ‘VTs’) indicates an apparent link between magma bulk viscosity and the occurrence of an approximately 90° horizontal rotation of accompanying VT earthquake fault-plane solutions (FPS), which may reflect stresses produced in the walls of an inflating or pressurizing dike. Furthermore, analysis of magmas from eruptions with and without 90° FPS rotations, provides a tight constraint on the minimum bulk viscosity associated with stress field reorientation. Finally, detailed reanalysis of the timing of normal faulting events during four representative eruptions of high bulk viscosity magma indicates that their presence or absence may be indicative of slow magma ascent (vs rapid ascent or stalling), providing a potential means to qualitatively assess magma ascent rates in near-real time.