Vagus nerve stimulation (VNS) is an effective alternative treatment for patients with refractory epilepsy. The generator produces intermittent stimulation trains and does not require patient intervention. Using currently available technology, continuous stimulation is incompatible with a reasonable battery life. Because earlier studies have demonstrated the persistence of a stimulation effect after discontinuation of the stimulation train, we intended to evaluate the clinical efficacy of VNS in both the programmed intermittent stimulation mode and the magnet stimulation mode. Patients, companions, and caregivers were instructed on how to administer additional stimulation trains when an aura or a seizure onset occurred. We assumed that patients or caregivers could recognize habitual seizures and were able to evaluate sudden interruption of these seizures. During a mean follow-up of 35 months, 46% of patients became responders, with a reduction in seizure frequency of more than 50%. Twenty-nine percent of patients stopped having convulsive seizures. In two thirds of patients who were able to self-administer or receive additional magnet stimulation, seizures could be interrupted consistently or occasionally. More than half of the patients who reported a positive effect of magnet stimulation became responders. Only three patients were able to use the magnet themselves. In most cases, support from caregivers was necessary. This study is the first to document the efficacy of magnet-induced VNS in a larger patient population during long-term follow-up. The magnet is a useful tool that provides patients who are treated with VNS and mainly caregivers of such patients with an additional means of controlling seizures. To further confirm the self-reported results from our patients, additional studies comparing programmed stimulation and magnet-induced stimulation during monitoring conditions are needed.