Background and Objectives: Patients requiring therapeutic diets in hospital are at risk of exposer to dietary er- rors that may pose an acute threat to their safety. This study aimed to determine the prevalence of meal-related er- rors among hospitalised patients prescribed therapeutic diets, following the implementation of an electronic food- service system (EFS). Methods and Study Design: This observational study involved six wards in a tertiary met- ropolitan hospital that used an EFS for meal ordering and plating. Participants were adult medical inpatients re- ceiving a therapeutic diet for medical or nutritional reasons. Meal accuracy was assessed for up to 48-hours per patient by comparing the dietary items placed on patients’ meal trays or personal meals consumed by patients to their therapeutic diet prescription. Inaccuracies were categorised as critical or non-critical errors and were identi- fied as having occurred at one of four steps in the EFS: menu planning (main-meals), meal assembly (main- meals), meal delivery (mid- and main-meals) and meal consumption (personal-meals). Results: A total of 167 in- patients were included in the study. Of the 906 meals assessed, 69 errors (8%) were observed; with 97% classi- fied as critical. Error rates differed according to the foodservice system step assessed: 17% for menu planning, <1% for meal assembly, 53% for meal delivery: main-meals, 9% for meal delivery: mid-meals and 33% meal consumption. Conclusion: An EFS almost completely eliminated errors associated with meal assembly. Howev- er, when foodservice staff and patients selected dietary items at ward level (without a guiding system) a substan- tial number of potentially critical errors occurred.