Verbally administered numerical rating scales (NRSs) from 0 to 10 are often used to measure pain, but they have not been validated in the emergency department (ED) setting. The authors wished to assess the comparability of the NRS and visual analog scale (VAS) as measures of acute pain, and to identify the minimum clinically significant difference in pain that could be detected on the NRS. This was a prospective cohort study of a convenience sample of adults presenting with acute pain to an urban ED. Patients verbally rated pain intensity as an integer from 0 to 10 (0 = no pain, 10 = worst possible pain), and marked a 10-cm horizontal VAS bounded by these descriptors. VAS and NRS data were obtained at presentation, 30 minutes later, and 60 minutes later. At 30 and 60 minutes, patients were asked whether their pain was "much less," "a little less," "about the same," "a little more," or "much more." Differences between consecutive pairs of measurements on the VAS and NRS obtained at 30-minute intervals were calculated for each of the five categories of pain descriptor. The association between VAS and NRS scores was expressed as a correlation coefficient. The VAS scores were regressed on the NRS scores in order to assess the equivalence of the measures. The mean changes associated with descriptors "a little less" or "a little more" were combined to define the minimum clinically significant difference in pain measured on the VAS and NRS. Of 108 patients entered, 103 provided data at 30 minutes and 86 at 60 minutes. NRS scores were strongly correlated to VAS scores at all time periods (r = 0.94, 95% CI = 0.93 to 0.95). The slope of the regression line was 1.01 (95% CI = 0.97 to 1.06) and the y-intercept was -0.34 (95% CI = -0.67 to -0.01). The minimum clinically significant difference in pain was 1.3 (95% CI = 1.0 to 1.5) on the NRS and 1.4 (95% CI = 1.1 to 1.7) on the VAS. The findings suggest that the verbally administered NRS can be substituted for the VAS in acute pain measurement.