Lacosamide (LCM) is a recently licensed antiepileptic drug available in the UK since 2008. It is thought to act through modulation of sodium channel slow inactivation. Its efficacy and tolerability have been shown in several regulatory randomised controlled trials, but assessments of its performance in large naturalistic settings are rare. We assessed a large cohort of consecutive people who started LCM at a single tertiary epilepsy centre, from June 2008 to June 2011. Forty-five percent of the 376 people included were still taking LCM at last follow-up, with estimated retention was 62% at one year, 45% at two years and 35% at three years. Eighteen percent reported a period of improvement in terms of significant seizure reduction or seizure freedom of at least six months duration whilst on LCM, of whom four people were seizure free for at least one year. Long-term efficacy in our centre appears similar to zonisamide and pregabalin when compared to historical controls. Adverse events were reported by 61%, CNS-related in the vast majority. Most clinical factors did not affect retention; withdrawal occurred more often because of inefficacy than because of adverse events. Retention rates for LCM, when compared to historical controls appear similar to lamotrigine, topiramate, pregabalin, zonisamide, higher than gabapentin, and lower than levetiracetam.