Purpose: The purposes of this study were to examine the relationship between health belief and exercise complianceamong elderly adults at senior centers and to identify factors influencing their exercise compliance. Methods: The subjects of this study were 100 elderly adults who were using senior centers in J City. Data were collected from the 5th of August to the 14th of September in 2014 using a questionnaire about general characteristics,health belief, and exercise compliance. Data analysis included one-way ANOVA, independent t-test, Pearson’s correlation, and stepwise multiple regression using the SPSS/WIN 18.0 program. Results: The mean score for exercise compliance was 3.85 (range 1~5), and for perceived health state 3.17 (range 1~5). The mean score for each of the sub-factors of health belief was 3.89 for benefit, 1.94 for barrier, 3.34 for severity, 2.43 forsensitivity, and 3.65 for exercise self-efficacy (range 1~5). There was a significant correlation between exercise compliance and exercise benefit, and 28% of variance in exercise compliance was explained by exercise benefit in health belief, family history of illnesses, and perceived sensitivity in health belief. Conclusion: To promote exercise compliance among elderly adults at senior centers, exercise programs emphasizing exercise benefit should be developed.