Should the declining diversity of weed communities in conventionally managed arable fields be regarded as a problem? The answer to this question has tended to divide researchers into those whose primary focus is on conserving farmland biodiversity and those whose goals are dictated by weed control and maximising yield. Here, we argue that, regardless of how weeds are perceived, there are common ecological principles that should underpin any approach to managing weed communities, and, based on these principles, increasing in-field weed diversity could be advantageous agronomically as well as environmentally. We hypothesise that a more diverse weed community will be less competitive, less prone to dominance by highly adapted, herbicide-resistant species and that the diversity of the weed seedbank will be indicative of the overall sustainability of the cropping system. Common to these hypotheses is the idea that the intensification of agriculture has been accompanied by a homogenisation of cropping systems and landscapes, accounting for both declines in weed diversity and the reduced resilience of cropping systems (including the build-up of herbicide resistance). As such, weed communities represent a useful indicator of the success of rediversifying systems at multiple scales, which will be a central component of making agriculture and weed control more sustainable.