We compared magnetic resonance imaging (MRI), magnetic resonance angiography, and transcranial Doppler ultrasonography as predictors of specific neurocognitive functions in children with sickle cell disease. Participants were 27 children with sickle cell anemia (hemoglobin SS) who were participants in the Stroke Prevention Trial in Sickle Cell Anemia (STOP) and had no documented history of stroke. Children's MRIs were classified as normal or silent infarct, and their magnetic resonance angiograms were classified as normal or abnormal. The highest time-averaged mean flow velocity on transcranial Doppler ultrasonographic examination of the major cerebral arteries was analyzed. Age and hematocrit also were analyzed as predictor variables. The battery of neurocognitive tests included measures of intellectual functioning, academic achievement, attention, memory, visual-motor integration, and executive functions. MRI, magnetic resonance angiography, transcranial Doppler ultrasonography, age, and hematocrit were analyzed as predictors of participants' performance on the various measures of neurocognitive functioning. Age and hematocrit were robust predictors of a number of global and specific neurocognitive functions. When age and hematocrit were controlled, transcranial Doppler ultrasonography was a significantly unique predictor of verbal memory. We found an association between low hemoglobin and neurocognitive impairment. We also found that abnormalities on transcranial Doppler ultrasonography can herald subtle neurocognitive deficits. (J Child Neurol 2006;21:37-44).