The use of Computational Fluid Dynamics (CFD) for process safety assessments in the Oil and Gas industry has become a widely accepted approach. Modern CFD codes can be used to simulate the progression of gas leaks in the near and farfield of a facility, predict the consequences of fire in terms of thermal fluxes (radiative and convective) on surfaces and the effects of smoke movement around a fire. Specialised gas dynamics codes, tuned to handle small time scale behaviours, can be used to predict overpressures from explosions. There is a vast array of work being performed to determine risks to personnel and structures, but how do we know what is modelled is a reasonable representation of the true scenario or at least a conservative one? Can we validate the work that is performed? If not, are we using best practice? Let us take consequence modelling of a gaseous release due to a leak or pipe failure as an example.