On 27 April 2016, White Island erupted in a multi-pulse, phreatic event that lasted for ~ 40 min. Six, variably sized pulses generated three ballistic ejections and at least one pyroclastic surge out of the inner crater and onto the main crater floor. Deposit mapping of the pyroclastic surge and directed ballistic ejecta, combined with numerical modelling, is used to constrain the volume of the ejecta and the energetics of the eruption. Vent locations and directionality of the eruption are constrained by the ballistic modelling, suggesting that the vent/s were angled towards the east. Using these data, a model is developed that comports with the field and geophysical data. One of the main factors modifying the dispersal of the eruption deposits is the inner crater wall, which is ~ 20 m high. This wall prevents some of the pyroclastic surge and ballistic ejecta from being deposited onto the main crater floor but also promotes significant inflation of the surge, generating a semi-buoyant plume that deposits ash high on the crater walls. While the eruption is small volume, the complexity determined from the deposits provides a case study with which to assess the relatively frequent hazards posed by active volcanoes that host hydrothermal systems. The deposits of this and similar eruptions are readily eroded, and for complete understanding of volcanic hazards, it is necessary to make observations and collect samples soon after these events.