To analyze differences in functional brain images collected in patients with type 2 and type 1 narcolepsy compared to normal controls and the relationship among brain images, objective neuropsychologic tests, and sleep findings. Data collection included comprehensive clinical investigation, study of sleep/wake with actigraphy, polysomnography, Multiple Sleep Latency Test, human leukocyte antigen typing, 18F-fluorodeoxyglucose PET, and cognitive tests obtained from 29 patients with type 2 narcolepsy, 104 patients with type 1 narcolepsy, and 26 sex- and age-matched normal control individuals. Conners' Continuous Performance Test (CPT II) and Wisconsin Card-Sorting Task were performed simultaneously with the FDG-PET examination. After analyses of variance, data between patients with type 1 and type 2 narcolepsy were compared by post hoc analysis, and correlation between functional brain imaging findings and results of neurocognitive tests was obtained. All patients with narcolepsy presented with at least 2 sleep-onset REM periods (SOREMP) and subjective sleepiness. Patients with type 2 narcolepsy compared to patients with type 1 narcolepsy had significantly less SOREMP, longer mean sleep latencies, and lower body mass indexes, apnea-hypopnea indexes, and frequency of human leukocyte antigen DQ-Beta1*0602. In patients with type 2 narcolepsy, FDG-PET studies showed significantly less hypermetabolism in the fusiform gyrus, striatum, hippocampus, thalamus, basal ganglia, and cerebellum than in patients with type 1 narcolepsy, and significantly less hypometabolism in the regions of frontal lobe, posterior cingulum, angular gyrus, and part of the parietal lobe; these changes were associated with fewer errors on the CPT. Young patients with type 2 narcolepsy have fewer clinical impairments and less distinct brain functional abnormalities than patients with type 1 narcolepsy, who are significantly more affected.