Containment is a frequently advocated strategic objective for countering plant invasions. It is commonly perceived that it is the valid fall-back option when eradication has failed or is deemed impossible with the available resources. We reviewed management and research literature on containment. The lack of a clear, universally accepted definition of containment is problematic and containment practice is not well aligned with the limited research literature. Vague and inconsistent use of the term, poorly developed relationships between management practice and the ecological drivers of invasion, and frequent failure to specify appropriately scaled spatial configurations in management strategies make it difficult to evaluate containment as a strategy, generally or in specific cases. Management strategies rarely provide the basis for effective and efficient containment programs and this may reflect a lack of under-pinning scientific principles. We recommend that containment be defined as 'deliberate action taken to prevent establishment and reproduction of a species beyond a predefined area' and suggest that containment efforts should focus on individual infestations or populations but simultaneously cover all infestations or populations that are separated from one another by habitat suitable for the species. Containment units should be rigorously defined but the inevitability of breaches of these containment units, due to the stochastic nature of dispersal, implies that containment generally requires a capacity for local eradication (extirpation). This means that many infestations are no more amenable to containment than to eradication. The measures we propose would improve assessment of the feasibility and success of containment strategies.