Weeds can cause significant problems to natural ecosystems. Although there have been numerous studies on the economics of weed control, relatively few of these studies have focused on natural ecosystems. This paper addresses this gap in the literature by assessing the cost-effectiveness of a comprehensive range of control strategies for blackberry (Rubus anglocandicans) in natural environments in Australia. We developed a stochastic dynamic simulation model and a deterministic dynamic optimisation model. The stochastic model calculates the expected net present value (NPV) of a range of control strategies, including any combination of treatment options. The optimisation model identifies the treatment combination that maximises NPV. Both models represent the costs and efficacies of control options over 25 years. The results indicate that using rust (Phragmidium violaceum) as a biological control agent only marginally increases NPV and excluding rust does not affect the optimal choice of other control options. The results also show for a wide range of parameter values that a strategy which combines the herbicide grazon (Triclopyre and picloram) and mowing is optimal. If chemical efficacy decreases by 20 percent it becomes optimal to include grazing blackberry by goats in the control strategy.