Globally, new combinations of introduced and native plant and animal species have changed rangelands into novel ecosystems. Whereas many rangeland stakeholders (people who use or have an interest in rangelands) view intentional species introductions to improve forage and control erosion as beneficial, others focus on unintended costs, such as increased fire risk, loss of rangeland biodiversity, and threats to conservation efforts, specifically in nature reserves and parks. These conflicting views challenge all rangeland stakeholders, especially those making decisions on how best to manage novel ecosystems. To formulate a conceptual framework for decision making, we examined a wide range of novel ecosystems, created by intentional and unintentional introductions of nonnative species and land-use–facilitated spread of native ones. This framework simply divides decision making into two types: 1) straightforward–certain, and 2) complex–uncertain. We argue that management decisions to retain novel ecosystems are certain when goods and services provided by the system far outweigh the costs of restoration, for example in the case of intensively managed Cenchrus pastures. Decisions to return novel ecosystems to natural systems are also certain when the value of the system is low and restoration is easy and inexpensive as in the case of biocontrol of Opuntia infestations. In contrast, decisions whether to retain or restore novel ecosystems become complex and uncertain in cases where benefits are low and costs of control are high as, for example, in the case of stopping the expansion of Prosopis and Juniperus into semiarid rangelands. Decisions to retain or restore novel ecosystems are also complex and uncertain when, for example, nonnative Eucalyptus trees expand along natural streams, negatively affecting biodiversity, but also providing timber and honey. When decision making is complex and uncertain, we suggest that rangeland managers utilize cost–benefit analyses and hold stakeholder workshops to resolve conflicts.