Background: Antibiotic prescribing for patients with urinary tract infections (UTIs) is often inconsistent with published guidelines. Pharmacists have a key role in the multidisciplinary team to improve the prescribing of antibiotics. Aim: This study determined the reasons for low guideline adherence and investigated the effects of a pharmacist-led, non-confrontational, inexpensive educational intervention on guideline adherence in adult inpatients with UTIs. Methods: The study was performed in three general medical wards in Auckland City Hospital (ACH), New Zealand, during March-September 2015. All patients treated for UTIs were identified by ward pharmacists. A pre-intervention observation period was followed by an intervention period in which ward pharmacists placed a simple diagnostic and treatment algorithm in the clinical records of all patients being treated for a UTI. Clinical and laboratory data available at the time initial antibiotic therapy was prescribed were used to determine a retrospective diagnosis for each patient. The initial antibiotic therapy was evaluated for adherence to the ACH antibiotic guidelines in relation to both the prescriber's and retrospective diagnosis. Results: Overall, the prescriber's diagnosis was concordant with the retrospective diagnosis in only 48 of 141 episodes (34%). Treatment was adherent to guidelines in 27% pre-intervention and 33% intervention episodes (p = 0.5). Conclusion: Inaccurate clinical diagnosis in patients with UTIs was a common reason for antibiotic treatment to be non-adherent to guidelines. The study failed to find an association between the educational intervention and an effect on guideline adherence. Passive interventions are unlikely to be effective and further interventions should consider codesign with users.UTIs: diagnostic accuracy and adherence to treatment guidelines c 'Other' diagnosis includes cellulitis, collapse, dehydration, shingles and upper respiratory tract infection.