We introduce the English-language Visual Numeric Scale (VNS) for self-reported pain and examine its psychometric properties; we compare the VNS to the better known Visual Analog Scale (VAS). We developed the VNS, which combines strong visual cues with an 11-point numeric rating scale. The VNS was administered to 2 sets of subjects with arthritis or chronic disease (N = 175, N = 192, respectively) and responses were examined. To compare the VNS to the VAS, we administered both scales to all subjects and used correlations to compare them to each other and to health distress and overall general health scores. A subset of respondents enrolled in an arthritis self-management program were given the VNS 4 months later, and change scores were used to test the sensitivity of the VNS. The VNS had means of 5.4 and 5.6 in the 2 samples, with distributions across the range of possible values. The VNS correlated well with the VAS (r = 0.85) and correlated slightly better than the VAS with the 2 independent health measures. The VNS was more likely to be completed than the VAS and there were fewer coding errors with the VNS. The VNS showed a significant (effect size 0.28) positive change for participants in a self-management course. The VNS appeared to be a valid measure. It was as successful as the VAS in measuring the underlying pain variable. It was easier to administer and code than the VAS, and was sensitive to change in pain.