OBJECTIVES: IBD-specific quality measures calling for the use of steroid sparing regimens (i.e., those involving anti-TNFs and non-biologic immunomodulators) were recently adopted by the Center for Medicare and Medicaid Services (CMS).  However, many older patients have contraindications to anti-TNFs.  Our objective was to describe drug utilization and patient characteristics among older IBD patients with anti-TNF contraindications.    METHODS:  A retrospective cohort study was conducted using CMS’ national 5% sample for 2006-2009 including Medicare patients with ≥12 months Parts A and B, ≥6 months Part D, an IBD diagnosis (≥2 claims for ICD-9CM 555.xx or 556.xx) and contraindications to anti-TNFs (advanced CHF, malignancies). We described the prevalence and days of exposure to each IBD drug class.  Patient characteristics associated with steroid exposure were examined using a negative binomial-logit hurdle model. RESULTS: Among 10,362 patients, 18% (n= 1860; 53% CHF, 39% malignancy, 8% both CHF and malignancy) had contraindications to anti-TNF therapy.  The mean age was 79 years, 67% were female and 87% white.  Steroid use ranged from 258-283 users per 1000 patients per year and averaged 123-145 mean annual treatment days for utilizers.  Anti-TNFs and non-biologic immunomodulators were used infrequently (anti-TNFs: 19-30 users per 1000 patients each year, non-biologic immunomodulators: 29-36 users per 1000 patients each year).  Patients who were younger, white, receiving any IBD drug class except anti-TNFs, had polypharmacy, more hospitalizations or absence of stroke history had greater odds of receiving steroids.  Among steroid recipients, polypharmacy and anti-TNF use were associated with a 65% (23%-121%) and 79% (28%- 150%) greater number of steroid therapy days, respectively.   CONCLUSIONS:  Use of steroids exceeded steroid sparing regimens supporting the importance of the new quality measure as a strategy to improve care.  Patients with anti-TNF contraindications are frequently appropriate to receive non-biologic immunomodulators and our findings suggest that these drugs are underutilized in this cohort.