The objective of this study was to determine the relationship between shoulder pain and the position of the seat of a wheelchair relative to the ground and to determine the relationship between shoulder pain and structural damage. A transversal study of a patient cohort of 140 patients with grade A and B spinal cord injuries below the T1 vertebra, who were full-time users of self-propelled wheelchairs, was made. The main variables are angle of the seat of the wheelchair, nuclear magnetic resonance measurements, shoulder joint range, Spinal Cord Index Measurement scale value, Wheelchair User Shoulder Pain Index and visual analogue scale value. Of the 140 patients, 99 were males (70.7%), and 41 were females (29.3%). Of these patients, 78 (55.7%) used a wheelchair seat at a straight angle, whereas 62 (44.3%) maintained an acute angle. A total of 57 (40.7%) patients presented shoulder pain within the last 6 months, and 83 (59.3%) did not. Sixty-nine (49.3%) did not present any type of shoulder injury as determined by nuclear magnetic resonance, whereas 71 (50.7%) had pathological signs. Patients with the seat at a straight angle developed 1.86 times (P<0.01) more shoulder pain and 1.73 times (P<0.001) more structural injuries than those with seats at an acute angle. Patients with structural abnormalities suffered 6.3-fold greater shoulder pain (P<0.001) than those without abnormalities. Wheelchair seats positioned parallel to the ground may be a risk factor for pain and shoulder injury. Patients with structural injuries in their shoulders have a greater risk of shoulder pain.