This paper examines the macroscopic consequences of wettability variations at the pore-scale. Important theoretical issues relating to distributed contact angles are examined and a process simulator capable of simulating the full flooding cycle characteristic of laboratory wettability tests is described. Simulations predict that final waterflood recoveries from weakly water-wet, or even weakly oil-wet, systems can often exceed those from strongly water-wet systems-even though initial imbibition rates may lead to the opposite conclusion. The underlying pore level physics has been explored by defining a capillarity surface which incorporates the combined effects of both contact angle and pore dimension during the imbibition process. Finally, a regime-based framework has been developed which can be used to reconcile apparently contradictory wettability experiments.