Engaging in regular physical activity that requires a moderate to vigorous effort is related to good health and a reduced risk of a multitude of diseases. Currently physical activity and health promotion research is limited by self- report data that is confounded by participants' ability to accurately remember and record their physical activity. A potentially new method of accurately assessing moderate-intensity physical activity is called context sensitive- ecological momentary assessment (CS-EMA). Mobile computing devices (personal computing devices, PDAs) that automatically identify when a specific type of physical activity (e.g., moderate-intensity physical activity) is being performed has the potential to provide valuable objective data for naturalistic, experimental, and clinical intervention research purposes, and has the potential for educational, recreational, and clinical applications. The goal of this exploratory study is to assess the accuracy of two methods of collecting data on moderate-intensity physical activity. The study will compare the use of a paper diary to collect self-reported data on moderate-intensity physical activities performed throughout the day to self-reported data collected using ecological momentary assessment methods which are facilitated by a PDA that cues the participant to record when specific intensity levels of physical activity are achieved. This study will also develop algorithms that will automatically identify specific types of physical activities being performed. The data collected from the heart rate monitor, motion sensor, and information requested directly from the participant via the PDA interface will be used to train pattern recognition systems to classify the intensity and type of physical activity being performed by the participant. This study will contribute to research on context- awareness and human activity recognition.