OBJECTIVES: The primary objective of this study is to describe self-reported work performance among a sample of employees with cancer. METHODS: This study uses data from the Truven Health Analytics Health and Productivity Management (HPM) and Commercial Claims and Encounters (CCAE) databases containing healthcare claims and self-reported performance. The study population consists of individuals over a 5-year period with at least one inpatient claim or two cancer outpatient claims more than 30 days apart with ICD-9 diagnosis codes for one of the following cancer types: breast cancer (female only), prostate cancer, colon and rectal cancer, melanoma, cancer of the lung and bronchus, or multiple myeloma. This abstract sub-samples 4,224 individuals from this cancer pool who completed at least one health risk appraisal containing self-reported performance outcomes. Socio-demographic and co-morbid characteristics are described and included in multivariate regression models. An additional multilevel analysis describes the degree of variation in performance that is attributed to employer characteristics as opposed to employee factors. RESULTS: Across all types of cancer individuals report non-zero work performance loss. Among the six cancer types individuals with breast cancer reported the highest levels of performance loss while those with prostate reported the lowest. Significant variation exists across employers in work performance. CONCLUSIONS: These types of work performance losses are usually not included in studies of medical costs, but they are important outcomes for employees, employers, families and society at large. Variation across employers and employees in these types of outcomes, beyond health care costs, warrants further attention. Access to high quality and appropriate treatment may mitigate the effects of work performance loss and prevent longer term periods of work disability and income disruption.