Standalone ‘low-cost’ inertial measurement units (IMUs) could facilitate large-scale studies into establishing minimal important differences (MID) for orthopaedic deficits (lameness) in horses. We investigated accuracy and limits of agreement (LoA) after correction of magnitude-dependent differences of a standalone 6 degree-of-freedom IMU compared with an established IMU-based gait analysis system (MTx) in six horses for two anatomical landmarks (sacrum and sternum). Established symmetry measures were calculated from vertical displacement: symmetry index (SI), difference between minima (MinDiff) and difference between maxima (MaxDiff). For the sacrum, LoA were ^0.095 for SI, ^6.6mm for MinDiff and ^4.3mm for MaxDiff. For the sternum, LoA values were ^0.088 for SI, 5.0mm for MinDiff and ^4.2mm for MaxDiff. Compared with reference data from mildly lame horses, SI values indicate sufficient precision, whereas MinDiff and MaxDiff values are less favourable. Future studies should investigate specific calibration and processing algorithms further improving standalone IMU performance.