A modified "Kluver" or dexterity board was developed to assess fine control of hand and digit movements by non-human primates during the acquisition of small food pellets from wells of different diameter. The primary advantages of the new device over those used previously include standardized positioning of target food pellets and controlled testing of each hand without the need for restraints, thereby allowing the monkey to move freely about the cage. Three-dimensional video analysis of hand motion was used to provide measures of reaching accuracy and grip aperture, as well as temporal measures of reach duration and food-pellet manipulation. We also present a validated performance score based on these measures, which serves as an indicator of successful food pellet retrieval. Tests in three monkeys show that the performance score is an effective measure with which to study fine motor control associated with learning and handedness. We also show that the device and performance scores are effective for differentiating the effects of localized injury to motor areas of the cerebral cortex.