There is now an extensive range of anti-epileptic drugs (AEDs) available including older established treatments and a newer generation of medications. The choice of drugs and what constitutes optimal therapy, however, is unclear due to limitations in the data supporting their use, particularly among the newer treatments. In clinical trials of monotherapy, a treatment is required to show only non-inferiority to another benchmark treatment. In trials of polytherapy, comparisons are limited to placebo. It is therefore necessary to look beyond the study data and consider other parameters to ascertain the most suitable treatment for the individual patient. Available evidence suggests that efficacy is similar among most AEDs, but this does not mean they are all the same. Some show efficacy in early and refractory epilepsy and some improve depression and quality of life (QOL) in epilepsy. AEDs are associated with a range of adverse events (AEs) that can limit their usefulness. AE classifications include type A (augmented and dose related) including tiredness, fatigue, insomnia, dizziness, vertigo, imbalance, ataxia, tremor and cognitive impairment; type B (bizarre and idiosyncratic) including various hypersensitivity reactions; type C (chronic long-term toxicity) including hirsutism, alopecia, weight gain and obesity; and type D (teratogenesis and carcinogenesis). The newer AEDs have been more thoroughly assessed for AEs than older drugs and risks are better understood. In AED safety, it is not better to follow a policy of ‘better the devil you know’ but rather to carefully monitor AE incidence and be prepared to switch drugs to improve tolerability and avoid non-compliance and treatment failure.