A nonexperimental design was used to determine whether the verbal scores of low-income gifted fifth graders (n = 38) differed from those of their higher income peers (n = 83). The Otis-Lennon School Ability Test, Eighth Edition and the Stanford Achievement Test-Tenth Edition were used to collect student data. Results of a MANOVA showed a statistically significant difference between the verbal scores of the two groups, with low-income students scoring significantly lower. A large effect size for the multivariate main effect of income level on verbal intelligence and verbal achievement scores was found (?[superscript 2] = 0.19). The existence of verbal-nonverbal score discrepancy in low-income students questions the practice of using only nonverbal or nonverbal parts of an IQ test to identify and place students in gifted programmes. These results also underscore the need to nurture underdeveloped verbal abilities when they occur in low-income students.