This study was undertaken to ascertain the prevalence of the persistent cavum septi pellucidi in children and adults by magnetic resonance imaging (MRI) and to compare the clinical indications for neuroimaging in the two age groups as a measure of group selection bias. All scans performed at West Virginia University during 1997 were reviewed for the presence of a persistent cavum septi pellucidi. The clinical indications for the MRI study were determined in 100 consecutive adult (17 years of age or older) and 100 consecutive pediatric (younger than 17 years of age) scans. In the 203 pediatric patients the prevalence of a persistent cavum septi pellucidi was 6.9%, and in the 814 adults the prevalence was 2.1%. Mental retardation/developmental delay was the clinical indication for at least 26% of the pediatric patients but was not an indication for neuroimaging in the adult study group. The known association of persistent cavum septi pellucidi with mental retardation and in several groups of patients with conditions clinically characterized by mental dysfunction suggests that the higher prevalence in the pediatric study group may primarily be the result of the patient selection bias operating through the different clinical indications for neuroimaging in the two populations.