Vesicular nanoparticles recovered from edible and medicinal plants have moved in little more than a decade from a curiosity of plant pathology to a proposed drug delivery platform. Their appeal is straightforward: grapefruit and ginger yield lipid material at the gram-per-kilogram scale, the starting biomass is agricultural rather than a bioreactor, and a carrier built from a food crop carries no risk of transmitting a human pathogen. Preclinical performance has matched the promise in places. Oral ginger and grape vesicles reach the inflamed colon and the intestinal crypt, folate-decorated grapefruit lipid particles have shown tumour accumulation more than a thousand-fold above unmodified controls, and ginseng vesicles have extended median survival in an orthotopic glioma model. This review synthesises the evidence across biogenesis, composition, isolation, cargo loading, surface engineering and therapeutic application, and argues that the field's principal obstacle is neither delivery efficiency nor toxicity but definitional. The entity most papers call a plant exosome is, in the majority of published work, a mechanically produced mixture of secreted vesicles, intracellular membranes and reassembled lipid particles, and the International Society for Extracellular Vesicles now advises against the term. Reported yields for a single species vary several-fold across laboratories, encapsulation efficiencies span 1% to 96% depending on the cargo, and dose has been reported in at least five mutually inconvertible units. Three registered clinical trials, the earliest opened in 2011, have between them enrolled fewer than one hundred participants and posted no results. The carrier is ahead of the analytics that would allow it to be manufactured, dosed and compared, and closing that gap is the precondition for translation.