Epidemiological studies have shown that low physical activity is associated with increased risk of several chronic diseases. The relationship between physical activity and health outcomes varies considerably, partly due to the difficulty of assessing physical activity accurately. Accurate measurement of physical activity is a pre-requisite to monitor population physical activity levels and design effective interventions. Accelerometer is an objective and precise technique to adequately assess physical activity and its association with health outcomes. However, there are still some issues related to accelerometer selection and usage. The purpose of this paper is to provide a review of the literature on the general principle, methodological decisions, and validity of accelerometers to assess physical activity.