Pterosaur wings bore a striking resemblance to sails, having a bony spar at the leading edge, formed by the forelimb and one enormously elongated digit, and an elastic wing membrane. Such simple wings would be expected to have performed badly due to excessive deformation, membrane flutter and poor control characteristics. Here I discuss how certain anatomical features, specifically a forewing membrane in the inner part of the wing and a system of fibres embedded in the distal part, may have countered these shortcomings. The forewing, supported by the unique pteroid bone, would have reduced the wings' geometric twist, and has been shown in wind tunnel tests to improve membrane stability at low angles of attack and dramatically increase the maximum lift coefficient at high angles of attack. The function of the fibres is poorly understood, but it is suggested that they improved membrane stability and optimised twist nearer the wingtips.