It has long been recognised that uncontrolled childhood epilepsy is detrimental to cognition and behaviour, impacting on a patient's ability to succeed academically. Patients who experience more frequent and serious seizures are at greater risk for cognitive decline, emphasising the need for more effective epilepsy treatments to bring seizures under control. That said, although more effective antiepileptic drugs (AEDs) have the potential to limit the impact of uncontrolled seizures on cognitive and behavioural function, recently it has been acknowledged that deficits in these functions may be caused by AEDs themselves. The cognitive and behavioural effects of older-generation AEDs have been determined largely from AE reporting rather than from specific assessment. Recently, clinical trials of newer-generation AEDs, such as topiramate, levetiracetam and perampanel, have included standardised neuropsychological tests as outcome measures to assess their impact on cognition and behaviour in children and adolescents. However, to understand how we may limit the cognitive and behavioural side effects of AEDs, it is necessary for us to gain a fuller, more accurate, characterisation of their true impact. Such insight will depend on sophisticated and standardised approaches to the design of AED clinical trials. This review provides a general overview of our current understanding of the impact of both epilepsy and AEDs on cognition and behaviour, before focusing on the AEDs for which more detailed assessment, using standardised cognitive and behavioural measures, has been undertaken. We will then go on to discuss the key elements in the design of future AED clinical trials to address current unmet needs.