The advancement of three-dimensional, time-dependent fire behaviour models is best supported by publicly available, co-located, synchronised, quality-assured measures of pre-fire, active fire and post-fire conditions (i.e. integrated datasets). Currently, there is a lack of such datasets. Consequently, we discuss essential components to produce integrated datasets: metadata, implementation of geospatial and temporal standards, data management plans, quality assurance project plans and data quality objectives. We present example data quality objectives and a data model for grassland experiments developed based on our experience integrating data from the 2014 Camp Swift Fire and the 2012 Prescribed Fire Combustion and Atmospheric Dynamics Research experiments.