Different sources of uncertainty in CFD simulations are illustrated by a detailed study of2-D, turbulent, transonic flow in a converging-diverging channel. Runs were performedwith the commercial CFD code GASP using different turbulence models, grid levels,and flux-limiters to see the effect of each on the CFD simulation uncertainties. Twoflow conditions were studied by changing the exit pressure ratio: the first is a complexcase with a strong shock and a separated flow region, the second is the weak shockcase with no separation. the uncertainty in CFD simulations has been studied in termsof five contributions: (1) iterative convergence error, (2) discretization error, (3) errorin geometry representation, (4) turbulence model, and (5) the downstream boundarycondition. in the discussion of the discretization error, results obtained from the RAE 2822 airfoil cases were also included as a representative study of the external flows.We show that for the simulation of attached flows, informed CFD users can obtainreasonably accurate results, whereas they are more likely to get large errors for the casesthat have strong shocks with substantial flow separation. We demonstrate the difficultyin separating the discretization errors from physical modelling uncertainties originatingfrom the use of different turbulence models in CFD problems that have strong shocks andshock-induced separation. For such problems, the interaction between different sourcesof uncertainty is strong, and highly refined grids, which would not be used in generalapplications, are required for spatial convergence. This study provides observations onCFD simulation uncertainties that may help the development of sophisticated methodsrequired for the characterization and the quantification of uncertainties associated withthe numerical simulation of complex turbulent separated flows.