Reasons for performing studyLameness negatively affects the welfare of horses; it often coexists with thoracolumbar pain. The mechanisms linking the two are not well understood.ObjectivesTo document thoracolumbar movement in subjectively sound horses comparing straight lines with circles on left and right reins; to relate these observations to the objectively determined symmetry/asymmetry of hindlimb gait.Study designCross‐sectional study.MethodsFourteen non‐lame horses were trotted in straight lines and lunged on a 10 m diameter circle on left and right reins and inertial sensor data collected at landmarks: withers, 13th (T13) and 18th thoracic vertebrae, 3rd lumbar vertebra, tubera sacrale, left and right tubera coxae. Data were processed using published methods [1]; pitch and roll, dorsoventral and lateral motion and symmetry within each stride were assessed.ResultsDorsoventral motion during one stride had a sinusoidal pattern with 2 oscillations. The greatest amplitude and symmetry (119 ± 4 mm, 96 ± 1% in straight lines vs. 127 ± 5 mm, 94 ± 1% in circles) occurred at T13. Circles induced greater rotation around the transverse axis (>1.3°, P = 0.002) and movement in a lateral direction (>16 mm P = 0.002), greater dorsoventral amplitude and asymmetry; however, the latter differences were nonsignificant. There were no significant differences between reins. The difference in the left and right hindlimb stance phases and the 2 oscillations of the thoracolumbar were significantly associated. Greater circle‐induced upward movement of a tuber coxae during the contralateral hindlimb stance was associated with greater circle‐induced asymmetry of the 2 oscillations of the thoracolumbar (P = 0.04).ConclusionsMoving on a circle induces measurable changes in thoracolumbar movement compared with moving in straight lines, associated with alterations in the hindlimb gait.Ethical animal research: The study was approved by the Ethical Review Committee of the Animal Health Trust. Owners gave informed consent for their horses' inclusion in the study. Source of funding: RVC PhD funding of Line Greve. Competing interests: None declared.