42 standardised 8 layer POP cylinders of appropriate dimensions were fabricated, of which 21 were split longitudinally. The splints were subjected to non-destructive tests in 4-point bending (Bending), 3-Point Kinking (kinking) and torsion modes, and the load at clinically relevant end-points was recorded. These simulated the deformity at which the splint no longer provided adequate stability and alignment, or at which the wearer was no longer protected. The splints were then loaded to destruction to establish the mode of ultimate failure. Results: The mean loads at the clinical end points for split POP splints were: 1375 N in Bending, 544 N in Kinking and 12 Nm in Torsion (equalling 67.3%, 70.4% and 47.4% of the equivalent values for a circumferential splints). Loads were in excess of body weight for most paediatric patients. After ultimate failure, the proportion of casts that became unstable was similar (44% of full casts and 50% of split casts). Conclusion: Split POP splints which have not been spread, provide adequate stabilisation and protection of paediatric forearm fractures, and do not routinely require completion