Significance/background Complementary and alternative medicine (CAM) therapies are often used with conventional medical treatments; hence healthcare providers need to understand factors influencing use and implications for patient care. Since estimates of use by cancer survivors vary widely (7–64%), population-level studies are needed to compare usage patterns. Purpose This study compares prevalence and patterns of CAM use in the US cancer population and the noncancer population. Factors that predict use of CAM providers, practices and products in relationship to symptom experience and purpose of use were identified. Methods A secondary analysis of the 2002 National Health Interview Survey data using STATA 9.2 software for population estimation was conducted. CAM use by cancer survivors ( N =2262; estimated cancer population=14.3 million) and individuals without cancer ( N =28,734; estimated population=191.5 million) was compared. Responses to 22 CAM use items in the NHIS Alternative Health Supplement were analyzed. Predisposing (gender, age, race, education, marital status), enabling (income, health insurance, provider-contact), and need factors (symptoms, co-morbidity, health status) in the CAM Healthcare Model were analyzed for ability to predict CAM use from a health service utilization perspective. Data analysis Binary and Multinomial Logistic Regression, the primary statistical models employed in the analysis, focused on between-subject differences in CAM use. A stepwise procedure was followed and potential predictor variables were excluded from the model if their p -value exceeded 0.10. Findings and implications Cancer and noncancer populations differ in their CAM use patterns, with cancer survivors more likely to use CAM. Predisposing, enabling and need factors were strong predictors of CAM use in both populations. Use was more prevalent among female, middle-aged, white, and well-educated people; women were specifically more likely to use CAM practices than men. A curvilinear relationship of use with age was noted suggesting a possible cohort effect. Higher income, private insurance, contact with conventional providers, pain and co-morbidity were strong predictors of CAM use. More than 39% of all cancer survivors used CAM, highlighting the need for oncology nurses and healthcare providers to assess use by their patients. Data will be compared with the soon to be released 2007 NHIS data set to determine trend patterns in CAM use over time.