The study was designed to evaluate optimal use of add-on lamotrigine in the treatment of children and adults with refractory epilepsy of any type. Because of the available evidence from controlled studies, indicating the large spectrum of action of lamotrigine, we designed this prospective study to investigate the efficacy and safety of lamotrigine in everyday clinical practice, to collect useful information on its action in specific epilepsy syndromes and on the clinical results of specific co-medications. We studied 566 patients with a diagnosis of refractory epilepsy of any type currently receiving stable conventional regimens of antiepileptic therapy. Efficacy analysis was limited to 510 patients (388 patients aged 12 years or more, 122 patients aged 2 to 12 years) for which the exact number of seizures could be evaluated. Seizure characteristics were: simple and/or complex partial seizures in 298 (58p. cent) patients, partial seizures with secondary generalisation in 85 (17p. cent), generalised seizures of any type in 226 (44 percent). Syndromic diagnosis was partial symptomatic or cryptogenic epilepsy in 302 patients (59 percent), generalised symptomatic or cryptogenic epilepsy in 116 (23 percent) and idiopathic generalised epilepsy in 50 (10 percent). The percentage of patients who achieved at least 50 percent reduction in the frequency of seizures was evaluated around 40p. cent for all epilepsy categories, and up to 61 percent in idiopathic generalised epilepsies. Response to treatment with lamotrigine was usually obtained by the end of titration (4 weeks) and remained stable at 48 weeks. Thirty-three patients (7 percent) remained seizure-free at 48 weeks. In the group of patients with partial epilepsy, 19p. cent presented a more than 75 percent reduction in seizure frequency. A more than 50 percent reduction in secondary generalisation of partial seizures was observed in 45 percent of the patients. Efficacy results were similar in both the adult and paediatric age groups. They were better for patients receiving valproate co-medication (45 percent of the responders) as compared to other co-medications (37 percent of the responders), suggesting a synergistic action. Safety has been evaluated for all the patients having received lamotrigine (n=566). The incidence of adverse events attributed to lamotrigine was similar to the results of controlled studies, with somnolence reported in 10p. cent, a cutaneous reaction in 8 percent and episodes of transitory diplopia in 8 percent. A cutaneous reaction was more frequent in patients receiving carbamazepine (10 percent) as compared to valproate comedication (5 percent). However, the adverse event was sufficiently serious to necessitate hospitalisation in 3 patients receiving valproate. Dose escalation was not respected in two. Rash was reversible in all of the patients following discontinuation of the drug. The results of this study contribute to the overall understanding of the spectrum of lamotrigine effectiveness across seizure types and epileptic syndromes. Lamotrigine was well tolerated in children and adults.