Needle injection through the outer annulus fibrosus of the intervertebral disc (IVD) is the most practical approach for delivery of therapeutic agents, which have been shown to potentially leak following needle retraction. The goal of this work was to establish a protocol for quantifying post-injection leakage and test its sensitivity to factors believed to affect needle track geometry. A through-puncture defect procedure, followed by controlled injection, was performed on bovine caudal IVDs. Sensitivity to needle size was tested by injection of saline into unconstrained discs with either a 30G, 26G, or 21G hypodermic needle. Sensitivity to axial load was tested by repeated injection via a 26G needle with either no constraint, fixed height, or 10% axial compressive strain. Sensitivity to flexion was tested by applying combined 0.2MPa compression and 15° of flexion following injection of 5% of disc volume. Needle diameter significantly affected maximum volume prior to leakage, ranging from 34.6 ± 31.9µL when using 21G to 115.6 ± 23.6µL when using 30G. While all unloaded discs leaked, axial compression decreased the incidence of leakage events by 50-100% depending on load history. Forward flexion resulted in a 22% incidence of leakage. Fluid injected into IVDs is at significant risk of leakage following needle retraction. This risk depends on factors which alter the geometry of the needle track, including needle size, pinching due to axial compression, and stretching as a result of forward flexion.