Millions of dollars and countless hours are spent on invasive plant management, and the field of invasion ecology has gained increasing attention in recent decades. Yet, despite these efforts to control and understand plant invasions, successful management is often elusive. Budgetary constraints are a common factor limiting invasive plant management programs, and therefore, optimizing control strategies are essential. However, such optimization requires data on management inputs and outcomes, and these data are often missing, lacking, or underutilized. To address this knowledge gap and identify predictors of successful weed control in natural areas, we examined nearly 20 years of invasive plant treatment data in the world’s largest urban national park – Santa Monica Mountains National Recreation Area of southern California. We resurveyed 279 sites that had undergone control in the past two decades, collecting data on the abundance of native and invasive plant species to evaluate long-term management outcomes. We used multiple statistical approaches to identify management inputs and site characteristics that are predictors of eradication, invasive plant cover, and native species recovery. We found that the greater the initial size or percent cover of an infestation, the lower the probability of eradication. We also found that weed infestations on steeper slopes and in areas that have burned more frequently are less likely to be eradicated. Promisingly, our results also showed that greater reductions in invasives generally benefited native plant communities, though not in all cases. These analyses also highlighted that persistence is key; more frequent treatments (both chemical and nonchemical) and greater investment of labor resulted in larger reductions in invasive plants. Our results highlight how site characteristics and limited resources can complicate invasive plant management, while demonstrating the value of analyzing treatment and monitoring data to identify effective control strategies and guide adaptive management decisions.