A new instrument was designed to provide a practical clinical measure for assessing children's pain intensity and pain affect. The pocket size measure includes a Coloured Analogue Scale (CAS) to assess intensity and a facial affective scale to assess the aversive component of pain. Both scales have numerical ratings on the back, so that the person administering it can quickly note the numbers that represent a child's pain. This study was conducted to determine the validity of the new instrument by evaluating the psychophysical properties of the intensity scale and by evaluating the discriminant validity of the intensity and affective scales. Since visual analogue scales (VAS) are valid and reliable measures for assessing children's pain, children's ability to use the new analog scale was compared with their performance on a VAS. Children's ability to rate pain affect using an affective scale, in which the 9 faces on a Facial Affective Scale (FAS) are presented in an ordered sequence from least to most distressed, was compared to their performance on the original FAS, in which the same faces were presented in a random order. Using a parallel groups design, 104 children (5-16 years; 60 female, 44 male; 51 healthy and 53 with recurrent headaches) were randomized into two groups: CAS or VAS. Children used the assigned scale to complete a calibration task, in which they rated the sizes of 7 circles varying in area (491, 804, 1385, 2923, 3848, 5675 and 7854 mm2). The psychophysical function relating perceived circle size to actual physical size was determined for the CAS and VAS. Children's CAS and VAS responses on the calibration task yielded similar mathematical relationships: psi cas = 0.035I0.87, psi vas = 0.027I0.89, where psi = perceived magnitude and I = stimulus intensity. The R2 values were 0.921 and 0.922 for the CAS and VAS groups, respectively. Analyses of covariance revealed no significant differences in the characteristics of these relationships, i.e., R2, slope, or y intercept, by scale type. Children used the same scale to complete the Children's Pain Inventory (CPI), in which they rated the intensity and affect of 16 painful events (varying in nature and extent of tissue damage). Children's CAS and VAS responses on the CPI were similar. Analyses of covariance indicated that there were no differences in either intensity or affective ratings by scale type. However, the mean number of painful events experienced by children increased significantly with age (P = 0.0001). Intensity ratings decreased significantly with age (P = 0.002), but affective ratings did not vary with age. The new instrument has equivalent psychometric properties to a 165 mm VAS. However, the CAS was rated as easier to administer and score than the VAS, so it may be more practical for routine clinical use. Since the CAS has fulfilled the first two criteria for a pain measure (psychophysical properties and discriminant validity), it is ethical to proceed with the formal definitive test for construct validity, in which children from various clinical populations use the CAS scale to assess their own pain.