OBJECTIVES: Clinical trials in epilepsy may not routinely prioritise patient-oriented outcomes that consider the harms of treatments in addition to their benefits.  To date, no systematic empirical research has been undertaken to assess the views of people with epilepsy about treatment outcomes.  The aim was to identify which outcomes of drug treatment are considered important to three groups of adults with epilepsy: (i) recently diagnosed, (ii) established diagnosis, (iii) women of childbearing age.   METHODS: Semi-structured individual interviews containing ranking exercise were used to explore views and interpretations of benefits, harms, and potential life-impacts of anti-epileptic drug treatments (n=41); the feasibility of these findings were evaluated in focus groups of health care professionals responsible for prescribing anti-epileptic drugs  (n=8) .  Outcomes ranked 1-4 were scored 4-1. For each group, scores were summed and divided by the number of participants.   RESULTS:  10 recently diagnosed men (mean age 45.9), 13 established (mean age 39.3, 92% male), and 18 women of childbearing age (mean age 34.5) participated.  Reduction in seizure frequency was the most highly ranked outcome of drug treatment across all three subgroups (women of childbearing age [score]=2.5, recent=2.4, established=2.23).  Adults recently diagnosed were most concerned about feelings of aggression (1.6), depression (1.0) and ability to work (0.9).  Adults with established epilepsy were most concerned with ability to work (1.15) negative impacts on relationships (1.0), memory problems (0.69), and social-life (0.69).  Women of childbearing age were concerned about memory (1.22), reduced independence (0.78), feeling in control (0.56) and foetal abnormality (0.5).  Clinicians considered life-impacts (eg work, relationships, independence) as consequences of benefits and harms of treatment.       CONCLUSIONS:  The importance of remission from seizure was consistent. However, patients’ rankings of unfavourable outcomes of drug-treatment varied by subgroup. Selection of outcome measures in clinical trials in epilepsy must consider relevant patient-oriented outcomes which differ by population.