The role of vegetable and fruit intake in reducing falls risk in elderly populations is uncertain. This study examined the associations of vegetable and fruit intake with falls-related hospitalisations in a prospective cohort study of elderly women (n 1429, ≥70 years), including effects on muscular function, which represented a potential causal pathway. Muscular function, measured using grip strength and timed-up-and-go (TUG), and vegetable and fruit intake, quantified using a validated FFQ, were assessed at baseline (1998). Incident falls-related hospitalisation over 14·5-year follow-up was captured by the Hospital Morbidity Data Collection, linked via the Western Australian Data Linkage System. Falls-related hospitalisation occurred in 568 (39·7 %) of women. In multivariable-adjusted models, falls-related hospitalisations were lower in participants consuming more vegetables (hazard ratio (HR) per 75 g serve: 0·90 (95 % CI 0·82, 0·99)), but not fruit intake (per 150 g serve: 1·03 (95 % CI 0·93, 1·14)). Only total cruciferous vegetable intake was inversely associated with falls-related hospitalisation (HR: per 20 g serve: 0·90 (95 % CI 0·83, 0·97)). Higher total vegetable intake was associated with lower odds for poor grip strength (OR: 0·87 (95 % CI 0·77, 0·97)) and slow TUG (OR: 0·88 (95 % CI 0·78, 0·99)). Including grip strength and TUG in the multivariable-adjusted model attenuated the association between total vegetable intake and falls-related hospitalisations. In conclusion, elderly women with higher total and cruciferous vegetable intake had lower injurious falls risk, which may be explained in a large part by better physical function. Falls reduction may be considered an additional benefit of higher vegetable intake in older women.