Context: Lower back pain is prevalent in horse riders as a result of the absorption of repetitive and multiplanar propulsive forces from the horse. Global positioning system technology provides potential for in vivo measurement of planar loading during riding. Objective: To quantify the uniaxial loading at the lumbar and cervicothoracic spine during dressage elements. Design: Repeated measures, randomized order. Setting: Equestrian arena. Patients (or Other Participants): Twenty-one female dressage riders. Intervention(s): Each rider completed walk, rising trot, sitting trot, and canter trials in a randomized order. A global positioning system unit was placed within customized garments at C7 and L5, collecting triaxial accelerometry data at 100 Hz. Outcome Measures: PlayerLoad based on the rate of change of acceleration and calculated in the anteroposterior (AP), mediolateral, and vertical planes during each trial. Results: There was no significant main effect for global positioning system location in the AP (P = .76), mediolateral (P = .88), or vertical (P = .76) planes. There was a significant main effect for pace in all trials (P < .001), with successive elements eliciting significantly greater loading (P ≤ .03) in all planes in the order walk < rising trot < canter < sitting trot. There was a significant placement × element interaction only in the AP plane (P = .03) with AP loading greater at L5 during walk, rising trot, and canter trials, but greater at C7 during sitting trot. Conclusions: The significant main effect for dressage element was indicative of greater pace of the horse, with faster pace activities eliciting greater loading in all planes. In vivo measurement of spinal accelerometry has application in the objective measurement and subsequent management of lumbar load for riders.