A control allocation architecture is developed for the control of the planar dynamics of a wheel motored vehicle. First, a benchmark controller is presented which takes advantage of indeterminate tire information to make informed control allocation decisions. This optimal controller is then simplified by an assumption of small slip-ratio, resulting in a controller which relies on only estimable parameters and is thus practical. Unfortunately, this suboptimal control allocation receives no information relating to tire saturation and its performance is significantly degraded compared with the optimal controller. In response, a holistic measure of tire saturation based only on estimable quantities is proposed and incorporated into the suboptimal control allocation routine. The result is a real-time implementable control system whose performance approaches that of the optimal benchmark control system.