Summary Intravenous fluid boluses guided by changes in stroke volume improve some outcomes after major surgery, but invasive measurments may limit use. From October 2016 to May 2018, we compared the agreement and trending ability of a photoplethysmographic device (Clearsight) with a PiCCO, calibrated by thermodilution, for haemodynamic variables in 20 adults undergoing major elective surgery. We analysed 4519 measurement pairs, including before and after 68 boluses of 250 ml crystalloid. The bias and precision of stroke volume measurement by Clearsight were −0.89 ± 4.78 ml compared with the invasive pulse‐contour cardiac output device. The coefficient of agreement for stroke volume variation after fluid boluses between the two devices was 0.79 (‘strong’). Fluid boluses that increased stroke volume by ≥ 10% increased mean absolute volume (SD) and mean percentage (SD) stroke volume measurements similarly for the invasive pulse‐contour cardiac output and Clearsight devices: 9 (4) ml vs. 8 (4) ml and 16% (8%) vs. 15% (10%), respectively, p > 0.05. The non‐invasive Clearsight pulse‐contour analysis was similar to an invasive pulse‐contour device in measuring absolute and changing stroke volumes during major surgery.