We report hole mobilities obtained computationally based on both single crystal geometries and those obtained from crystal fragments optimised on a model surface. Such computational estimates can differ considerably from experimentally measured thin film mobilities. One source of this discrepancy is due to a difference in the morphology of the thin film compared with that of the crystal. Here, predictions of thin film hole mobilities based on optimised structures are given. A model surface is used to provide an inert geometric platform for the formation of an organic monolayer. The model is tested on pentacene and TIPS-pentacene for which experimental information of the surface morphology exists. The model has also been applied to four previously uninvestigated structures. Two of the compounds studied had fairly low predicted mobilities in their single crystal structures, which were vastly improved post-optimisation. This is in accord with experiment.