Research suggests that children suffering from different types of disorders (learning disorders, behavioral disorders, or intellectual disabilities) are sometimes evaluated differently simply due to the presence of a diagnostic label. We conducted a multilevel meta-analysis of experimental studies (based on data from 8,295 participants and on 284 effects nested in 60 experiments) to examine the magnitude and robustness of such label effects and to explore the impact of potential moderators (type of evaluation, diagnostic category, expertise, student's gender, and amount and type of information). We found a moderately negative overall label effect (Hedges' g = -0.42), which was robust across several types of evaluation, different samples, and different diagnostic categories. There was no indication that expertise and the gender of the child moderated the effect. Presenting participants with only a label yielded the strongest negative effect of g = -1.26, suggesting that the effect was dependent on the amount of information being presented to participants. We conclude that labeling a child can exacerbate negative academic evaluations, behavioral evaluations, evaluations of personality, and overall assessments of the child. Further implications for theory and future research are discussed.