This thesis describes in detail the clinical spectrum of spinal stenosis in a series of two hundred and twenty-one patients at the Nuffield Orthopaedic Centre. It depicts those conditions with which spinal stenosis may be confused, and other conditions with which it is associated. Characteristic symptoms and physical signs are reported and the role and value of different methods of investigation are discussed. The aetiology and pathogenesis of spinal stenosis is discussed and the emphasis turned away from absolute measurements of the dimensions of the bony spinal canal, towards the role of the soft tissues and the dynamic response of the canal and its neural contents to postural change and loading, as evidenced by erect flexion and extension radiculography. The spinal reserve capacity measurement on CT approaches more closely the ideal of measurement of volumetric disproportion of canal and contents, but it takes no account of the dynamics of the canal. Magnetic Resonance Imaging may, in the future, provide the most objective criteria for diagnosis if section thickness can be reduced. Experimental spinal stenosis was produced in a group of immature New Zealand white rabbits. This was induced by sublaminar wiring at three levels at the age of eight weeks and allowing the animals to grow for twenty-four months before sacrifice and analysis of the spines. The effect of sublaminar wiring on the growth and development of the lamina and spinal canal was analysed using a Kontron Ibas Image Analysis Computer, and the results described and statistically analysed. The results of surgery were analysed in detail in a group of seventytwo patients with spinal stenosis at the Nuffield Orthopaedic Centre. The long-term results were compared with the initial post-operative result and two groups were identified: the stable result and the unstable result. The indications for and results of re-operation were also analysed in a group of twelve patients. Improved understanding of the aetiology of spinal stenosis has enhanced surgical management and results. The extent of surgical decompression must be precisely planned pre-operatively from radiographic and CT studies, and the surgeon must be able to execute this plan at operation. There is now no place for exploratory operations. The objective of surgery is adequate nerve root decompression without spinal de-stabilisation and when this is achieved, re-operation is redundant.