Aims: To develop an antibiotic foot formulary for the empirical treatment of diabetes-related foot infections presenting to our service. Subsequently, to asses costs associated with the introduction of our protocol, in particular to assess the effect on admissions avoidance and any cost savings achieved. Methods: We reviewed several existing antibiotic protocols. We analysed data on costs related to treatment and admission rates prior to and after the introduction of the protocol. Results: We rationalized our antibiotic protocol and adapted the Infectious Disease Society of America guideline by introducing a category of 'moderate infection-borderline admission' to our classification. This enabled the administration of outpatient intramuscular antibiotics. After introducing the rationalized protocol, our average antibiotic prescribing costs for a 3-week course of treatment fell from £17.12 to £16.42. Over 22 months of follow-up, 26 episodes were eligible for treatment with intramuscular antibiotics. Over the same time period, 121 people were admitted directly from the foot clinic. The costs saved as a result of avoided or delayed admission for those 26 episodes was over £76 000. For 12 people who required subsequent admission, their length of hospital stay was significantly shorter than those admitted directly [9.25 days (range 2-25) vs. 16.11 (2-64), P = 0.045]. Conclusions: By modifying the Infectious Disease Society of America classification and adopting a protocol to administer outpatient oral and intramuscular antibiotics, we have led to substantial cost savings, shorter hospital admissions and also have developed a successful admissions avoidance strategy.