OBJECTIVES: Individual data sources contain non-integrated data components needed to assess outcomes, resource use, and costs in cancer patients. This work describes methodology to integrate disparate electronic data sources in chronic myelogenous leukemia (CML) patients with a common identifier (CI). METHODS: A CML Patient cohort from the Huntsman Cancer Institute was created by extracting information across the Utah Cancer Registry; the Utah Population Database (UPDB); and the Enterprise Data Warehouse, including Cerner inpatient and EPIC ambulatory care clinic data. Medication use was from inpatient medication orders. A unique patient index identifier linked disparate records. RESULTS: A total of 602 patients were identified by ICD-9 diagnosis code for CML (250.1, 205.10-12) from 1995 through 2009, median age = 51, 42.6% female. Of these 598 (99.3%) were linked to the UPDB and 245 had a state death certificate. Charlson Comorbidity Index (CCI) analysis (+/- 90 days) identified 232 (38.5%) subjects with a score of zero, 199 (33.1%) with 1-3, 99 (16.4%) with 4-6, 47 (7.8%) with 7-9 and 25 (4.2%) with a score of 10-17 (median=2, mean= 2.6, and SD= 3.1). Inpatient admission data was available for 380 (63.1%) patients, with a total of 267 CML related drug orders. Procedures were observed for 531 (88.2%) patients. Lab results were available for 564 (93.7%) subjects. Of those, BCR/ABL biomarker results were available for 210 (37.2% of all lab results) patients. CONCLUSIONS:   Integrating data across different data sources in an academic health care center with a National Comprehensive Cancer Network hospital can provide comprehensive health care data. This methodology may influence the evolution of electronic health records, as a data resource tool for outcomes data, resource use and cost utilization across complex disease states such as CML. Future research will expand on drug data sourcing and evaluate the medical record notes to evaluate CML specific outcomes.