INTRODUCTION: Autonomic instability is considered a contributing factor in sudden unexpected death in epilepsy (SUDEP). The aim of this pilot study was to measure parasympathetic activity in sub-clinical seizures to investigate autonomic instability. MATERIALS AND METHODS: A prospective study based on Video-electroencephalography (EEG)/electrocardiography (ECG)/oxygen saturation (SAO2) recordings was selected from patients having sub-clinical seizures during stage 3 or 4 sleep. We analysed R-R intervals in the ECG from 1-min prior to the electrographic onset to the end of sub-clinical seizures. Matched non-ictal R-R baseline measurements were selected from stages 3 or 4 sleep. R-R interval data were analysed using NeuroScope software providing a cardiac index of parasympathetic activity (CIPA). BioSignal short-term heart rate variability (HRV) software was used to analyse the same R-R interval data previously analysed using NeuroScope except that sub-clinical seizure data was embedded within 5-min epochs and compared to 5-min epochs of non-ictal measurements. RESULTS: A total of 33 sub-clinical seizures were recorded from 11 patients comprising 19 generalised sub-clinical seizures (2 patients), 9 right temporal lobe sub-clinical seizures (5 patients) and 5 left temporal lobe sub-clinical seizures (4 patients) were compared to matched non-ictal measurements. Parasympathetic activity was clearly altered during total sub-clinical seizures in terms of the CIPA (p<0.001) and 5-min HRV high frequency (HF) % (p=0.026) measures. Generalised sub-clinical seizures resulted in increased cardiac parasympathetic activity whereas temporal lobe seizures were associated with a decrease in parasympathetic activity. CONCLUSION: This pilot study indicates that parasympathetic changes occur during sub-clinical seizures. Generalised sub-clinical seizures may be associated with more autonomic instability compared to temporal lobe sub-clinical seizures.