Nesfatin-1 and ghrelin are the two recently discovered peptide hormones involved in the control of appetite. Besides its main appetite-control function, ghrelin also has anticonvulsant effects, while nesfatin-1 causes depolarization in the paraventricular nucleus (PVN). The aims of this study, therefore, were to investigate: (i) whether there are differences in the concentrations of nesfatin-1 and ghrelin in saliva and serum samples between eplilepsy patients and normal controls and (ii) whether salivary glands produce nesfatin-1. The study included a total of 73 subjects: 8 patients who were newly diagnosed with primary generalized seizures and had recently started antiepileptic drug therapy; 21 who had primary generalized seizures and were continuing with established antiepileptic drug therapy; 24 who had partial seizures (simple: n = 12 or complex: n = 12) and were continuing with established antiepileptic drug therapy; and 20 controls. Salivary gland tissue samples were analyzed for nesfatin-1 expression by immunochemistry and ELISA. Saliva and serum ghrelin levels were measured by ELISA and RIA, and nesfatin-1 levels by ELISA. Nesfatin-1 immunoreactivity was detected in the striated and interlobular parts of the salivary glands and the ducts. The nesfatin-1 level in the brain was around 12 times higher than in the salivary gland. Before antiepileptic treatment, both saliva and serum nesfatin-1 levels were around 160-fold higher in patients who are newly diagnosed with primary generalized epilepsy (PGE) than in controls; these levels decreased with treatment but remained about 10 times higher than the control values. Saliva and serum nesfatin-1 levels from patients with PGE and partial epilepsies who were continuing antiepileptic drugs were also 10-fold higher than control values. Serum and saliva ghrelin levels were significantly (twofold) lower in epileptic patients before treatment than in controls; they recovered somewhat with treatment but remained below the control values. These results suggest that the low ghrelin and especially the dramatically elevated nesfatin-1 levels might contribute to the pathophyisology of epilepsy. Therefore, serum and saliva ghrelin and especially the remarkably increased nesfatin-1 might be candidate biomarkers for the diagnosis of epilepsy and for monitoring the response to anti-epileptic treatment.