Edge plasma modelling is discussed with particular emphasis on the comparison of a fluid neutral model with a kinetic neutral model. By iterative inclusion of additional effects, the agreement between the two is improved. Of particular importance is the proper treatment of neutrals at the core boundary, of the proper implementation of a neutral flux limit, and the modification of the ion heat boundary condition to include the neutral contribution. In the end, the agreement in the upstream profiles of electron and ion temperatures, and of electron and core neutral densities is very satisfactory. In addition, the effects of parallel ion and electron heat flux limiters, of transport ballooning to the low field side, and of drifts are investigated.