Microseismic monitoring is regularly used in the mining and geotechnical industries to provide detailed information about rock mass response to operational activities. This paper implements a variable 3D velocity model (VM) at a North American hard rock mine with the aim of further improving the accuracy of seismic event locations. A collapsing method is also used to help analyse two individual seismic event clusters at another North American hard rock mine. In the first case study, the location accuracy of events is shown to improve by moving from a conventional single VM to a variable 3D VM that accounts for voids and different geological domains. The modal average for the event location accuracy improved from 14 m to 5 m when the 3D VM was used. Additionally, the single VM failed to locate 26 per cent of the seismic events that were located by the 3D VM. In the second case study, the distribution of seismic events was further enhanced by using a collapsing method that highlighted planar structures. The collapsing method was performed on two large event sequences that occurred in close proximity and identified different seismic event distributions. One aftershock sequence aligned along a plane, suggesting an activated structure, while the other large event sequence had a more 3D distribution of events, suggesting a complex stress redistribution on smaller structures.CITATION:Pinnock, I, Collins, D S, Toya, Y and Hosseini, Z, 2016. Large event sequence analysis and 3D velocity models for seismic event location accuracy, in Proceedings Seventh International Conference and Exhibition on Mass Mining (MassMin 2016), pp 879-884 (The Australasian Institute of Mining and Metallurgy: Melbourne).