Background: Inappropriate antibiotic treatments for urinary tract infections (UTIs) in nursing homes cause the development of resistant bacteria. Nonspecific symptoms and asymptomatic bacteriuria are drivers of overtreatment. Nursing home staff provide general practice with information about ailing residents; therefore, their knowledge and communication skills influence prescribing. This paper describes the development of a tailored, complex intervention for a cluster-randomised trial that targets the knowledge of UTI and communication skills in nursing home staff to reduce antibiotic prescriptions. Methods: A dialogue tool was drafted, drawing on participatory observations in nursing homes, interviews with stakeholders, and a survey in general practice. The tool was tailored through a five-phase process that included stakeholders. Finally, the tool and a case-based educational session were tested in a pilot study. Results: The main barriers were that complex patients were evaluated by healthcare staff with limited knowledge about disease and clinical reasoning; findings reported to general practice were insignificant and included vague descriptions; there was evidence of previous opinion bias; nonspecific symptoms were interpreted as UTI; intuitive reasoning led to the inappropriate suspicion of UTI. Conclusion: Sustainable change in antibiotic-prescribing behaviour in nursing homes requires a change in nursing home staff’s beliefs about and management of UTIs.