Background: Neuroimaging studies of early-onset bipolar disorder (BD) are important in order to establish a fuller understanding of the underlying pathophysiology of the illness. The advantages of studying BD in children and adolescents include the relative absence of some confounds present in adult-onset research, such as lengthy duration of illness and exposure to treatments, greater number of mood episodes, and the presence of substance abuse or dependence. Finally, studying youths with the disorder may enhance our knowledge about the neural mechanisms of affective dysregulation and may specifically elucidate whether there are abnormalities that are unique to the early-onset form of the illness. Methods: PubMed was used to identify peer-reviewed publications from the past 15 years (January 1990 to January 2005) that used brain-imaging techniques (anatomic, functional, and biochemical) to research early-onset BD. Results: Eleven studies using anatomic magnetic resonance imaging (MRI), seven using magnetic resonance spectroscopy (MRS), and two using functional MRI (fMRI) were identified. Structural abnormalities were reported in total cerebral, white matter, superior temporal gyrus, putamen, thalamus, amygdala, and hippocampal volumes. Deficits in cortical gray matter were also reported. Using MRS, abnormalities were reported in the dorsolateral prefrontal cortex, anterior cingulate, and basal ganglia. One fMRI study found increased activation in the putamen and thalamus of BD youths compared to controls, and a second found abnormal prefrontal-subcortical activation in familial pediatric BD. Conclusion: Published MRI studies of early-onset BD are few. Nonetheless, extant data implicate abnormalities in brain regions thought to regulate mood and cognition. Synthesis of the findings into an overall model of anatomic and functional disruption is difficult due to the methodological variations among studies and the limitations of individual studies, such as the use of small sample sizes, the heterogeneity of sample characteristics, and the wide range of brain structures selected for analysis. Recommendations are offered to guide future research. It will be important for future studies to reproduce prior findings and determine which findings are unique to early-onset BD, relative to adult-onset illness. In addition, studies will need to establish the extent to which early-onset BD may overlap with comorbid disruptive, mood, anxiety, or psychotic disorders. (HARV REV PSYCHIATRY 2005;13:125-140.)