Background: Defining the reliability of oscillometry (NIBP) for measuring mean arterial pressure (MAP) during normotension and hypovolemic hypotension could help with cardiovascular monitoring in horses anaesthetised in dorsal recumbency. Objectives: To assess equivalence between NIBP and invasive MAP during normotension and hypovolemic hypotension in dorsally recumbent, anaesthetised horses. Study Design: Prospective experimental study using 9 cardiovascularly healthy mature horses. Methods: Under isoflurane anaesthesia, blood was removed every 10 minutes at 5 ml/kg bwt increments up to 25 ml/kg bwt and returned. Paired NIBP and invasive MAP measurements (facial, transverse facial, and metatarsal arteries; FA, TFA, MT) were obtained prior to and during blood removal/return and tested for equivalence using linear mixed-effects model. Sensitivity and specificity of NIBP to detect hypotension (MAP < 70 mmHg) were calculated, and an optimum NIBP MAP cut-off for hypotension was estimated. Results: A total of 257 paired NIBP and invasive MAP were compared. NIBP was equivalent to invasive MAP measured at the FA ( p = 0.3; M = -1, 95% CI [-3, 0.5]), with 88% sensitivity and 73% specificity to detect hypotension (defined as MAP < 70 mmHg in the FA). NIBP and MAP measurements from the FA remained equivalent when a standard 27 mmHg heart level correction was used ( p = 0.4; M = 1, 95% CI [-1, 3]). NIBP MAP < 80 mmHg using a heart level correction of 27 mmHg (or MAP < 107 mmHg without correction) is the optimum value defining hypotension (best balance between sensitivity, 95%, and specificity, 65%). Main limitations: Only adult horses in dorsal recumbency were studied. Conclusions: NIBP, as studied here, was equivalent to invasive MAP in the FA under normotension and hypovolemic hypotension. A NIBP MAP < 80 mmHg after 27 mmHg heart level correction appears to be the optimum value defining hypotension.