An observational study identified living people with epilepsy who matched (by age and geographical region) adolescents and adults with a postmortem diagnosis of sudden unexpected death in epilepsy (SUDEP). Use of an intervention to prevent SUDEP was determined for each participant who died (before entry into the study or subsequently). Fewer people died when an intervention to reduce SUDEP was employed (on average, 213 fewer deaths per 1000 people with nocturnal supervision and 184 fewer deaths per 1000 people with special precautions). However, data were available for only 367 of the 616 live participants recruited, and evidence was of very low quality, making firm conclusions impossible.