AbstractQuestionHow does urbanization and associated declines in fire frequency alter the floristic composition of native temperate grasslands? Does it lead to: (1) biotic homogenization, i.e. compositional similarity between remnants increases; (2) biotic differentiation, whereby similarity between remnants declines, or; (3) clustered differentiation, where similarity between remnants remains unchanged, but composition shifts from the historical state?LocationVictoria, Australia.MethodsUsing site‐level surveys, we examined changes in the floristic similarity of 29 urban grasslands from 1992 to 2013 and compared these changes to those of 63 rural grasslands from 1989 to 2014. Community‐level changes in the representation of key functional traits were also examined in urban grasslands, with traits advantaged following disturbance regime change and urban fragmentation predicted to increase in frequency.ResultsOur results supported the biotic homogenization hypothesis in urban grasslands. Compositional similarity between grasslands increased principally because of an increase in commonly shared non‐native species, with change in native composition comparatively minor. However, no evidence of biotic homogenization was found in rural grasslands, with no significant change in overall composition identified. The most urbanized sites had the highest number of non‐native species in both the current and historical data sets, yet non‐native composition over the past two decades changed the most in sites on the urban fringe, becoming more similar to sites closer to the urban core. As expected, following declines in fire frequency and increased urbanization, the overall composition of urban grasslands shifted to taller plant species, while native species capable of vegetative reproduction and exotic species with an annual life span increased in frequency.ConclusionUrbanization was an important driver of biodiversity change in the investigated system, with increasing competition intensity in response to disturbance regime change a likely cause of biotic homogenization. Our results demonstrate that non‐native species are a key driver of biotic homogenization, emphasizing the importance of managing non‐native immigration and maintaining historical disturbance processes once native ecosystems become urbanized.