To address the controversy regarding the relationship between cognitive impairment (lowering of IQ) and magnetic resonance imaging (MRI) characteristics (T2-weighted hyperintensities or unidentified bright objects [UBOs]) in children with neurofibromatosis-1 (NF-1), we used a pairwise NF-1/ sibling design; we set out to predict the lowering of IQ in each child with NF-1 as a discrepancy from the IQ of an unaffected sibling (D-SIQ). Our multiple regression model included the age of the child with NF-1, familial or sporadic nature of the NF-1, number of locations in the child's brain occupied by T2-weighted hyperintensities (UBOs), and the volumetric percentage of brain tissue occupied by T2-weighted hyperintensities (UBOs). Only the number of locations occupied by UBOs accounted for IQ lowering (D-SIQ) in children with NF-1 (42% of the variance in D-SIQ). This is the first report to confirm that a continuum of lowered IQs in NF-1-affected children exists in relation to the distribution of UBOs (range 0-7), not just presence (vs. absence) of any UBOs.