The term green island was first used to describe an area of living, green tissue surrounding a site of infection by an obligately biotrophic fungal pathogen, differentiated from neighbouring yellowing, senescent tissue. However, it has now been used to describe symptoms formed in response to necrotrophic fungal pathogens, virus infection and infestation by certain insects. In leaves infected by obligate biotrophs such as rust and powdery mildew pathogens, green islands are areas where senescence is retarded, photosynthetic activity is maintained and polyamines accumulate. We propose such areas, in which both host and pathogen cells are alive, be termed green bionissia. By contrast, we propose that green areas associated with leaf damage caused by toxins produced by necrotrophic fungal pathogens be termed green necronissia. A range of biotrophic/hemibiotrophic fungi and leaf-mining insects produce cytokinins and it has been suggested that this cytokinin secretion may be responsible for the green island formation. Indeed, localised cytokinin accumulation may be a common mechanism responsible for green island formation in interactions of plants with biotrophic fungi, viruses and insects. Models have been developed to study if green island formation is pathogen-mediated or host-mediated. They suggest that green bionissia on leaves infected by biotrophic fungal pathogens represent zones of host tissue, altered physiologically to allow the pathogen maximum access to nutrients early in the interaction, thus supporting early sporulation and increasing pathogen fitness. They lead to the suggestion that green islands are ‘red herrings’, representing no more than the consequence of the infection process and discrete changes in leaf senescence.