Introduction: Transverse ultrasound measurements of the median nerve (MN) for diagnosis of carpal tunnel syndrome (CTS) suffer from inconsistent findings within and between patients and healthy subjects. The objective of this study was to improve ultrasound assessment of CTS. Methods: In a case-control study (51 patients, 25 controls) we evaluated the performance gained by: (1) correcting for ultrasound probe angulation; (2) including active parameters such as forceful gripping of the hand; and (3) including hand flexor tendon parameters. Results: Correcting ultrasound probe angulation increased the correct classification rate by 4.5%; including forceful gripping resulted in increasing it by 2.8%; and including the hand flexor tendon resulted in an increase of 1.3%. Conclusions: The best predictive model combines correcting probe angulation with forceful gripping parameters and hand flexor tendon parameters. However, the clinically most practical model might use only probe angulation correction.