Drylands support over 2 billion people and are major providers of critical ecosystemgoods and services across the globe. Drylands, however, are one of the most susceptible bio-mes to degradation. International programmes widely recognize dryland restoration as key tocombating global dryland degradation and ensuring future global sustainability. While theneed to restore drylands is widely recognized and large amounts of resources are allocated tothese activities, rates of restoration success remain overwhelmingly low. Advances in understanding the ecology of dryland systems have not yielded proportionaladvances in our ability to restore these systems. To accelerate progress in dryland restoration,we argue for moving the field of restoration ecology beyond conceptual frameworks of eco-system dynamics and towards quantitative, predictive systems models that capture the proba-bilistic nature of ecosystem response to management.To do this, we first provide an overview of conceptual dryland restoration frameworks.We then describe how quantitative systems framework can advance and improve conceptualrestoration frameworks, resulting in a greater ability to forecast restoration outcomes andevaluate economic efficiency and decision-making. Lastly, using a case study from the westernUnited States, we show how a systems approach can be integrated with and used to advancecurrent conceptual frameworks of dryland restoration. Synthesis and applications.Systems models for restoration do not replace conceptual mod-els but complement and extend these modelling approaches by enhancing our ability to solverestoration problems and forecast outcomes under changing conditions. Such forecasting offuture outcomes is necessary to monetize restoration benefits and cost and to maximize eco-nomic benefit of limited restoration dollars.