In this article, we propose a novel anti-windup (AW) framework for coping with input saturation in the disturbance rejection problem of stable plant systems. This framework is based on the one developed by Weston and Postlethwaite (W&P) (Weston, P.F., and Postlethwaite, I. (2000), 'Linear Conditioning for Systems Containing Saturating Actuators', Automatica, 36, 1347-1354). The new AW-design improves the disturbance rejection performance over the design framework usually suggested for the coprime-factorisation based W&P-approach. Performance improvement is achieved by explicitly incorporating a transfer function, which represents the effect of the disturbance on the nonlinear loop, into the AW compensator synthesis. An extra degree of freedom is exploited for the coprime factorisation, resulting in an implicitly computed multivariable algebraic loop for the AW-implementation. Suggestions are made to overcome the algebraic loop problem via explicit computation. Furthermore, paralleling the results of former work (Turner, M.C., Herrmann, G., and Postlethwaite, I. (2007), 'Incorporating Robustness Requirements into Antiwindup Design', IEEE Transactions on Automatic Control, 52, 1842-1855), the additive plant uncertainty is incorporated into the AW compensator synthesis, by using a novel augmentation for the disturbance rejection problem. In this new framework, it is shown that the internal model control (IMC) scheme is optimally robust, as was the case in Turner, Herrmann, and Postlethwaite (2007) and Zheng and Morari (Zheng, A., and Morari, M. (1994), 'Anti-windup using Internal Model Control', International Journal of Control, 60, 1015-1024). The new AW approach is applied to the control of dynamically substructured systems (DSS) subject to external excitation signals and actuator limits. The benefit of this approach is demonstrated in the simulations for a small-scale building mass damper DSS and a quasimotorcycle DSS.