Instruction in the use of forceps and needle driver to suture tissues typically involves group didactic and demonstration sessions, followed by limited individual observation and coaching. Most motor learning required for dexterous suturing takes place during unsupervised practice with practice boards or during actual procedures in the operating room. We are developing surgical instruments with embedded microelectromechanical sensors for tracking instrument motion. Motion data is acquired and processed on a computer for concurrent or summary performance feedback during practice. An integral feature in our approach is the use of digital video recordings, synchronized with the sensor signals, to parse surgical procedures into a series of actions based on a task analysis. We envision the parsed video and signals as tools for assessment and performance feedback that an instructor could use to offer more extended individualized coaching. In a small pilot study, we concentrated our data analysis on the orientation of a needle driver about its long axis, the range of motion in one throw, and the timing of subtasks.