This study was designed to measure the validity of two variations of a self-report pain intensity scale. Sixteen children aged 9-17, suffering from migraine headaches and one parent of each child recorded headache intensity daily at fixed intervals for four weeks in headache diaries. Half of the children used a pain intensity scale with behavioral descriptors of pain severity accompanied by numbers zero to five; the remainder used a scale that was identical except for the addition of subjective pain descriptors. Validity was determined by assessing the degree of inter-rater concordance between the child's rating of intensity of head pain and the parent's independent rating of the same phenomenon. Concordance values, determined by weighted kappa, ranged from slight to almost perfect and all but one were significant, r<.05, providing evidence for the validity of this scale. Percentage agreement and weighted percentage agreement were also calculated and indicated high levels of concordance. Inclusion of subjective descriptors did not significantly affect ratings.