AD/HD is one of the most common psychiatric disorders of childhood, affecting approximately 5% of primary school children. Almost all models of the disorder accept that the behavioural cluster which is AD/HD results from an underlying central nervous system (CNS) dysfunction. However, the exact nature of this dysfunction is poorly understood. Several electrophysiological based models of AD/HD have been proposed and recent research has suggested that most are too simplistic in nature, and the underlying CNS dysfunctions are inaccurately labeled. Part of the problem results from the use of multiple bands in the analysis of the EEG, as this approach does not allow an understanding of the role of any discrete band on functioning. In a different approach, our group has been decomposing the EEG into single bands and relating anomalies in these bands to specific brain states (such as arousal), and to behavior. Results from a number of studies, and their implications for understanding the link between brain and behaviour, will be discussed.