Invasive blood pressure measurement (IBP) using peripheral arteries is a commonly used technique in equine anaesthesia, although the accuracy has not been demonstrated. Non-invasive blood pressure monitoring (NIBP) may be indicated for field anesthesia, short procedures and foal anaesthesia. In the present report, the agreement of various IBP and NIBP measuring sites compared to carotid artery pressure was tested in anaesthetised experimental ponies. Six ponies were anaesthetised in lateral recumbency with sevoflurane and received either saline or dexmedetomidine boli followed by constant rate infusion (CRI). Invasive blood pressure measurements were obtained simultaneously from the carotid, facial and metatarsal arteries. NIBP measurements over both median arteries, metatarsal and middle coccygeal arteries were performed in random order. All blood pressure readings obtained were compared to carotid pressure by Bland-Altman analysis. Non-invasive blood pressure measurements had larger bias and poorer limits of agreement compared to IBP measurements. NIBP measurement from the coccygeal artery had the best repeatability and best limits of agreement of all NIBP positions and was not affected by the use of dexmedetomidine. The facial artery had smaller limits of agreement with dexmedetomidine treatment. There was no difference between the facial and metatarsal arteries during dexmedetomidine treatment in MAP and DAP. Systolic arterial pressures from the metatarsal artery showed larger bias and larger limits of agreement compared to facial artery. In conclusion, NIBP measurement from the tail artery is a reliable alternative to direct arterial blood pressure measurement. Both the metatarsal and facial arteries have acceptable agreement with carotid artery pressure so the choice can be based on the logistics of the procedure.