Abstract We present a broad overview of the 2012–13 flank fissure eruption of Plosky Tolbachik Volcano in the central Kamchatka Peninsula. The eruption lasted more than nine months and produced approximately 0.55 km3 {DRE} (volume recalculated to a density of 2.8 g/cm3) of basaltic trachyandesite magma. The 2012–13 eruption of Tolbachik is one of the most voluminous historical eruptions of mafic magma at subduction related volcanoes globally, and it is the second largest at Kamchatka. The eruption was preceded by five months of elevated seismicity and ground inflation, both of which peaked a day before the eruption commenced on 27 November 2012. The batch of high-Al magma ascended from depths of 5–10 km; its apical part contained 54–55 wt. SiO2, and the main body 52–53 wt. SiO2. The eruption started by the opening of a 6 km-long radial fissure on the southwestern slope of the volcano that fed multi-vent phreatomagmatic and magmatic explosive activity, as well as intensive effusion of lava with an initial discharge of > 440 m3/s. After 10 days the eruption continued only at the lower part of the fissure, where explosive and effusive activity of Hawaiian–Strombolian type occurred from a lava pond in the crater of the main growing scoria cone. The discharge rate for the nine month long, effusion-dominated eruption gradually declined from 140 to 18 m3/s and formed a compound lava field with a total area of ~ 36 km2; the effusive activity evolved from high-discharge channel-fed 'a'a lavas to dominantly low-discharge tube-fed pahoehoe lavas. On 23 August, the effusion of lava ceased and the intra-crater lava pond drained. Weak Strombolian-type explosions continued for several more days on the crater bottom until the end of the eruption around 5 September 2013. Based on a broad array of new data collected during this eruption, we develop a model for the magma storage and transport system of Plosky Tolbachik that links the storage zones of the two main genetically related magma types of the volcano (high-Al and high-Mg basalts) with the clusters of local seismicity. The model explains why precursory seismicity and dynamics of the 2012–13 eruption was drastically different from those of the previous eruption of the volcano in 1975–76.