OBJECTIVES: The objective of this study was to determine the economic impact of a pharmacy program to convert insulin utilization from multi-dose vials to pen delivery systems on a long term care facility chain. METHODS: Purchasing data was obtained at the patient level for basal and short acting insulins from a chain of 75 skilled nursing facilities for the 12 month period ending June 2010. Data included date dispensed, amount dispensed (mls), delivery system (pen or vial) and amount paid to the dispensing pharmacy.  The insulin cost per patient-day for each month was calculated as total acquisition cost for the month divided by the number of patient-days.  The insulin cost per patient-day for each stay was calculated as the total insulin acquisition cost divided by the length of stay in days.  The mean cost per patient-day for each patient stay subset based on payer type, length of stay and delivery system used (pen only, vial only, pen and vial combination) was calculated.  RESULTS: There were 2,405 inpatient stays over the 12 month period, 70% covered by Medicare and 29% by Managed Care.  Two-thirds of Medicare stays and over three-fourths of managed care stays were 30 days or less.  Pen device purchases increased from under 1% to almost 35% of total purchases over the study period during which the insulin cost per day declined from over $10 per patient-day to $4.  The cost per day for vial-only stays ($7.84) and combination vial and pen stays ($7.79) were 72% higher than pen-only stays ($4.54), despite a 39% price premium per milliliter for pens.  Differences were most marked for lengths of stay under 30 days. CONCLUSIONS: The increase in pen device use was associated with a marked decrease in insulin costs on a patient-day basis, particularly for lengths of stay under 30 days.