Background: It is widely appreciated that vancomycin use is excessive. A drug use evaluation of intravenous vancomycin orders at our hospital showed that 48% were inappropriate. This finding was surprising because our stewardship program required Infectious Disease (ID) approval for all vancomycin orders, although it allowed an emergency dose prior to approval. Methods: After a comprehensive, hospital-wide educational program, all vancomycin orders were required to have ID pre-approval without exception. Two months after the new pre-approval process began, vancomycin orders were again evaluated for appropriateness. Results: After the intervention, there was a decrease in inappropriate prescribing from 48% to 22% of patients (p < 0.001), and a decrease in inappropriate days-on-therapy from 37% to 18% (p < 0.001). These improvements were due mainly to decreased prescribing of first dose, empiric use, in the Emergency Department (ED). No change was seen in clinical diagnosis, correctness of weight-based dosing, patient outcomes, or the rationale for vancomycin use. Common misconceptions amongst prescribers contributed to the overprescribing of vancomycin: (1) routinely adding vancomycin for empiric treatment of sepsis without risk factors for MRSA; (2) the false perception that vancomycin is necessary to “cover” various species of streptococci; (3) the need to treat a single positive blood culture growing coagulase-negative staphylococci; and (4) imprecision in assigning a diagnosis of “penicillin allergy”. Conclusions: Eliminating the provision for single dose vancomycin administration without ID pre-approval resulted in more appropriate vancomycin prescribing, particularly in the ED. Education to improve prescribing habits and dispel misconceptions about the empiric role of vancomycin was enabled by the pre-approval process.