The last decade has seen a major increase in use of the discrete fracture network (DFN) approach applied to cave mining problems, often integrated within more complex geomechanical simulations. Because the value of the DFN model directly depends on the quality and quantity of available field data, there is the need to apply dedicated mapping procedures to collect sufficient structural data at the required engineering scale. This paper attempts to address several key questions about DFN modelling; for example: are current methods for fracture mapping suitable for DFN modelling? What are the critical parameters that we need to measure and what do we need to change in the current practice? Particularly at the earlier stages of a cave project, engineers may have to account for the lack of appropriate data, which manifests itself in the form of increased uncertainty and the difficulty to characterise variability in an objective manner using a limited set of data. DFN models could in principle be generated using limited data by assuming statistical distributions for the sampled parameters; however, engineers would need to be aware of the implications that such limited knowledge (and increased uncertainty) would have on stability analysis, fragmentation analysis and mine design. When used correctly, a DFN approach offers practitioners the opportunity to:maximise the use of fracture data collected from mapping of rock exposuresgenerate more realistic synthetic rock mass modelsprovide estimates of natural fragmentation that also account for rock mass variability.CITATION:Elmo, D, Eberhardt, E, Stead, D and Rogers, S, 2016. Discrete fracture network modelling - importance of accounting for data uncertainty and variability, in Proceedings Seventh International Conference and Exhibition on Mass Mining (MassMin 2016), pp 845-852 (The Australasian Institute of Mining and Metallurgy: Melbourne).