The development and standardisation of an inventory for measuring anti-epileptic drug effects on cognition and affect is described. The Side Effect and Life Satisfaction inventory (SEALS) was derived from symptoms and side-effects reported by a patient population. It was administered to 45 patients on two occasions and test-retest reliability was demonstrated. It was administered to 923 patients with epilepsy. An underlying factor structure was produced by principal components analysis, consisting of five sub-scales which were stable when data from males and females were analysed separately or together. Validity was evaluated by comparing inventory scores of patients undergoing a number of treatment trials to show that the inventory was sensitive to treatment differences when other variables were controlled. Patients taking two or more anti-epileptic drugs had poorer SEALS scores than those taking a single drug. Patients taking vigabatrin and one other AED had poorer scores than those taking lamotrigine (LTG) and one other AED. A comparison of changes from baseline to week 4 of a double blind comparison of carbamazepine (CBZ) and LTG showed significantly greater improvement in SEALS scores for patients taking LTG, and significantly poorer scores in those patients taking CBZ who dropped out of the study, than in those who continued. It is concluded that SEALS is a valid and reliable tool for use in anti-epileptic drug trials.