Overuse injuries in youth baseball players due to throwing are at an all-time high. Traditional methods of tracking player throwing load only count in-game pitches and therefore leave many throws unaccounted for. Miniature wearable inertial sensors can be used to capture motion data outside of the lab in a field setting. The objective of this study was to develop a protocol and algorithms to detect throws and classify throw intensity in youth baseball athletes using a single, upper arm-mounted inertial sensor. Eleven participants from a youth baseball team were recruited to participate in the study. Each participant was given an inertial measurement unit (IMU) and was instructed to wear the sensor during any baseball activity for the duration of a summer season of baseball. A throw identification algorithm was developed using data from a controlled data collection trial. In this report, we present the throw identification algorithm used to identify over 17,000 throws during the 2-month duration of the study. Data from a second controlled experiment were used to build a support vector machine model to classify throw intensity. Using this classification algorithm, throws from all participants were classified as being “low,” “medium,” or “high” intensity. The results demonstrate that there is value in using sensors to count every throw an athlete makes when assessing throwing load, not just in-game pitches.