Objective. To examine the validity of a durometer to objectively measure skin hardness in systemic sclerosis (SSc), and to compare digital durometry with the modified Rodnan skin score (MRSS) and ultrasonography. Methods. Patients with SSc and healthy controls underwent durometry measurements in 3 assessments: a Latin square experiment to establish durometry's intra- and interobserver reliability compared with skin scoring (5 SSc, 1 control); a longitudinal cohort to assess sensitivity to change in skin hardness (13 SSc, 5 controls); and an ultrasound cohort to evaluate correlation between durometry, ultrasound-measured skin thickness, and clinical skin scoring (30 SSc, 12 controls). Results. Intraobserver reproducibility was higher for durometry than for clinical skin scoring (intraclass correlation coefficient [ICC] 0.97 versus 0.85), whereas interobserver reproducibility was similar (0.75 versus 0.73). Interobserver reproducibility of durometry was good for all body areas (ICC 0.61-0.85), but for skin scoring it was moderate in the legs (0.51) and poor in the abdomen (0.08), feet (0.09), and fingers (0.27). Durometry scores correlated with clinical skin scores (Latin square: r = 0.44, P = 0.03; longitudinal cohort: r = 0.81, P < 0.001) and ultrasound-measured skin thickness (hands: r = 0,58, forearms: r = 0.63, upper arms: r = 0,40; P ≤ 0.001 for all). Uninvolved skin in patients with SSc was harder than skin from controls (mean ± SD 23 ± 7 durometer units [DU] versus 19 ± 6 DU; P < 0.0001). Finally, there was a strong correlation between change in MRSS and change in durometry score (r = 0.77, P = 0.002). Conclusion. Durometer-measured skin hardness correlates well with MRSS and ultrasound-measured skin thickness, provides greater reliability than MRSS, and is sensitive to changes in skin hardness over time. Durometry should be considered for use in clinical therapeutic SSc trials. © 2006, American College of Rheumatology.