OBJECTIVES: Hypomethylating agents (HMAs), decitabine and azacitidine, are indicated for use in treatment of myelodysplastic syndromes (MDS), however only a minority of patients receive HMAs.  Our objective was to examine patterns of treatment associated with FDA-approved 5-day decitabine (DEC-5) and 7-day azacitidine (AZA-7) and off-label 5-day azacitidine (AZA-5) in MDS patients.   METHODS:  We identified MDS patients with an initial HMA treatment between July 1, 2005 to June 30, 2009 in 2 large insurance claims databases.  Index date was the date of initial HMA treatment. Patients were stratified into: DEC-5, AZA-7, or AZA-5, based on their first cycle of treatment and were followed for 6 months. We described the number of unique cycles of index treatment and treatment gaps (days of missed treatment) in these groups.    RESULTS: We identified 18,706 patients with MDS; 546 were treated with HMAs and were included in the study (156 received DEC-5, 176 received AZA-5 and 214 received AZA-7). Mean age was similar across groups: 68.8-71.2 years. Neutropenia was more common before treatment initiation in the DEC-5 (34.6%) group than in AZA-5 (22.7%) and AZA-7 (26.6%; p<.05) groups. There were 1,701 treatment cycles: 431 DEC-5 (per patient mean:2.8; median:2), 586 for AZA-5 (mean:3.3; median:3), and 684 for AZA-7 (mean:3.2; median:3) (p<0.05 for means).  DEC-5 cycles had the fewest gaps: 94.9% had no treatment gaps, compared to 89.1% for AZA-5 and 23.4% for AZA-7.  Among DEC-5 cycles, 3.2% had a 2 day gap, compared to 7.2% for AZA-5 and 66.5% for AZA-7 (p<0.001). CONCLUSIONS:   In this retrospective claims analysis, few MDS patients were treated with HMAs. Among those who received HMAs, decitabine patients were more likely to have prior neutropenia. Between the 2 FDA-approved regimens, DEC-5 and AZA-7, there were significantly fewer gaps with decitabine treatment. More treatment gaps were observed with use of longer AZA regimen.