A lifting-surface method is used to investigate the aerodynamic performance of two high aspect-ratio wings flying in formation. In particular, the influence of wake rollup is investigated. The relaxed-wake cases have performance results that differ from those obtained with a fixed-wake model. In particular, the fixed-wake model predicts more optimistic results for the follower aircraft for formations having a lateral separation of 70- to 90-percent span, which are also the lateral formation spacings with the greatest performance gains for the trailing aircraft. For formations with streamwise separations of a halfspan or more, the choice of wake model has very little influence on the changes resulting from the formation moving further apart in the streamwise direction. This reflects the fact that, in the case of the relaxed wake, the majority of the rollup of the lead-aircraft wake occurs within a short distance aft of its trailing edge and any further wake-rollup has little effect. In contrast, the fixed wake model experiences no changes at all after being shed. Increasing lateral and vertical separations reduce the differences due to a fixed versus relaxed wake. For the investigation of the performance of the formation, both aircraft are trimmed in pitch and roll. To a limited degree, the control inputs required to trim the formation depend on whether the relaxed- or fixed wake model is used, especially when the performance of the trailing aircraft is at an optimum.