Background: Poison Center (PC) data is often looked at for surveillance purposes as a means for identifying clusters of cases or unusual presentations. We present a case series of a patient with 17 separate emergency department (ED) encounters across 3 states for self-reported brake fluid ingestion. The PC was able to identify this patient as a "frequent flier" by his name and unique ingestion. By identifying the patient and recognizing a pattern of benign clinical outcome we averted administration of costly treatments. Case: Over 10 years (2003 . 2012) the PC documented 17 encounters, encompassing 3 states and 14 hospitals, of a man with self-reported brake fluid ingestion. Quantity of brake fluid ingested varied from unknown amount up to 12 ounces, with an average time of 13.8 hours (range 15 min . 30 hrs, n 13) before presentation to the ED. Mean initial serum bicarbonate was 24 mmol/L (range 21.6 . 27 mmol/L, n 15) and mean pH was 7.3 (range 7.33 .7.45, n 8). Osmolal gap was obtained on 3 cases (range 6 . 10 mOsm/L) and anion gap was calculated for 5 cases (median 12, range 11 . 24 mmol/L). He was prophylactically treated with fomepizole twice and ethanol infusion once after recommendations by a medical toxicologist. Final serum bicarbonate was documented in 7 cases with a mean of 24.6 mmol/L (range 21.5 . 29 mmol/L) and no elevation in serum creatinine was noted. Patient routinely complained of abdominal pain and requested opioids. Once this patient and pattern of ingestion was identified by the PC, further recommendations were for observation and monitoring of metabolic panel rather than empiric treatment. Case discussion: Brake fluid often contains diethylene glycol; ingestion can be fatal and may require costly interventions such as fomepizole or dialysis. This case series illustrates how our PC was able to provide management recommendations for a patient with potential toxic ingestion, suspected drug-seeking behavior, and no objective proof of significant toxicity based upon multiple prior encounters. By averting the use of empiric fomepizole in favor of observation, supportive care, and serial chemistries, the PC potentially saved these hospitals thousands of dollars. Conclusion: PCs should be cognizant of patients with multiple case records and attempt to track this data as it might guide future management recommendations. This case series emphasizes a unique opportunity for PCs to develop surveillance systems which may mitigate unnecessary healthcare spending.