BACKGROUND: Estimation of physical activity contributes to early screening and prevention of decrease in body function. This study was conducted to examine the reliability and validity of the International Physical Activity Questionnaire (IPAQ) short form in elderly people. METHODS: A total of 55 outpatients of 65 years old or older participated in this study. Physical activity over the last seven days was recorded on the IPAQ short form. Responses were converted to metabolic equivalent task minutes per week. The IPAQ short form was validated against accelerometer measurements as a gold standard. The test-retest method was performed over a two-week interval to examine the reliability of the IPAQ. RESULTS: The more physically active by the IPAQ short form, the higher the measured value of the accelerometer (P < 0.001). Pearson's correlation coefficient was 0.43 for the correlation between the results of two measurements. Spearman Rho coefficients and Kappa values of test-retest reliability in five elements (vigorous days, vigorous minutes, moderate days, moderate minutes, and walk days) were 0.299-0.605 and 0.307-0.418, respectively. CONCLUSION: The validity of the IPAQ short form was proven, but the reliability was found to be low. Nevertheless, IPAQ short form seems worthwhile tool for the measurement of physical activity levels, concerning daily variation in physical activity of the elderly.